semantica/CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

• Fix: KGVisualizer now accepts KnowledgeGraph objects in all visualize_* methods (PR visualization by @KaifAhmad1, closes #458): All five public methods (visualize_network, visualize_communities, visualize_centrality, visualize_entity_types, visualize_relationship_matrix) previously called graph.get("entities", []), silently producing no output when passed a non-dict object. Added _normalize_graph() which duck-types the input — dicts pass through unchanged; any object exposing .entities / .relationships attributes (e.g. the result of GraphBuilder.build()) is converted to the canonical dict form; anything else raises a clear ProcessingError naming the offending type. 21 tests added in tests/visualization/test_kg_visualizer_normalize_graph.py.

• Security: 12 vulnerability fixes across CRITICAL → LOW severity (PR security-enhancement by @KaifAhmad1):

  Critical

  • Eval injection eliminated (semantica/parse/media_parser.py): Replaced eval(stream.get("r_frame_rate", ...)) — which executed arbitrary Python from ffprobe JSON output — with a _safe_parse_fps() helper using fractions.Fraction. No code execution possible regardless of ffprobe output content. (CWE-95)
  • Unsafe pickle deserialization replaced (semantica/context/agent_memory.py): AgentMemory.save() / load() previously used pickle.dump / pickle.load, allowing RCE if an attacker could write the .pkl file. Replaced with json.dump / json.load. MemoryItem.to_dict() / MemoryItem.from_dict() added for safe round-trip serialization — timestamp via isoformat(), embedding dropped (not JSON-safe, regenerated on demand). Legacy .pkl files are detected and refused with a migration message. (CWE-502)

  High

  • SQL injection hardened (semantica/ingest/snowflake_ingestor.py): WHERE, ORDER BY, LIMIT, and OFFSET were f-string interpolated directly into Snowflake queries. LIMIT / OFFSET now use parameterized %s placeholders; ORDER BY is validated against a strict ^[A-Za-z_][A-Za-z0-9_]*(\s+(ASC|DESC))? regex; WHERE clauses containing semicolons are rejected before execution. (CWE-89)
  • XXE protection added for RDF/XML parsing (semantica/explorer/utils/rdf_parser.py): rdflib.Graph().parse() on XML-based RDF formats had no external-entity restrictions. Added _safe_parse_rdf() wrapper that calls defusedxml.defuse_stdlib() before parsing, neutralising Billion Laughs and local file-read XXE attacks. Falls back gracefully with a UserWarning if defusedxml is not installed. (CWE-611)
  • Security headers and CORS added to main server (semantica/server.py): No CORS policy, no response security headers, and raw Exception details were returned to clients. Added CORSMiddleware (origins from SEMANTICA_CORS_ORIGINS env var, defaults to localhost only); _SecurityHeadersMiddleware emitting X-Content-Type-Options, X-Frame-Options, X-XSS-Protection, Referrer-Policy, Permissions-Policy, and HSTS (HTTPS only) on every response; global error handler that logs internally and returns a generic 500. (CWE-346, CWE-200)
  • CORS and WebSocket hardened in Explorer app (semantica/explorer/app.py): allow_methods=["*"] and allow_headers=["*"] narrowed to GET, POST, DELETE, OPTIONS and Content-Type, Authorization only. KeyError / ValueError exception handlers now log the real message server-side and return generic text to clients. WebSocket messages larger than 64 KB trigger close with code 1009 (Message Too Big), preventing memory exhaustion via large frame injection. (CWE-346, CWE-400)

  Medium

  • Algorithm parameter validated by enum (semantica/explorer/routes/graph.py): The algorithm query parameter previously accepted any string; unknown values silently fell back to BFS. Replaced with _PathAlgorithm(str, Enum) — FastAPI now returns 422 Unprocessable Entity for any value other than bfs or dijkstra. (CWE-20)
  • RDF upload extension allowlist (semantica/explorer/routes/vocabulary.py): No file extension check was performed before reading RDF uploads. Extension is now validated against {".ttl", ".rdf", ".owl", ".xml", ".jsonld", ".json-ld", ".json"} before any content is read. (CWE-434)
  • Prompt injection mitigated (semantica/semantic_extract/llm_extraction.py): User-supplied text and entity/relation labels were embedded directly into LLM prompts via f-string interpolation — a crafted input like "\n\nIgnore all above instructions..." could override system instructions. All user-supplied content is now passed through json.dumps() before embedding, neutralising newlines, quotes, and instruction-override attempts. (CWE-1336)
  • Dynamic __import__() removed (semantica/pipeline/pipeline_validator.py): __import__("collections").Counter(...) replaced with a proper from collections import Counter module-level import. (CWE-95)
  • ReDoS eliminated (semantica/explorer/routes/enrich.py): re.split(r"\s+AND\s+", antecedent_text, re.IGNORECASE) on user-supplied rule strings could exhibit polynomial backtracking. Fixed by normalising whitespace first with " ".join(text.split()) (no regex) then splitting on the literal " AND ". Closes CodeQL alert #12. (CWE-1333)
  • Path traversal blocked (semantica/server.py): SPA catch-all route used STATIC_DIR / full_path without validation. Added Path.resolve() + relative_to() check that returns 400 Bad Request for any path that escapes STATIC_DIR. Closes CodeQL alerts #13 and #14. (CWE-22)

  Low

  • SPARQL result cap and timeout (semantica/explorer/routes/sparql.py): SPARQL queries ran to completion with no row limit or timeout; expensive queries could exhaust memory or block indefinitely. Results are now capped at 5 000 rows; asyncio.wait_for(..., timeout=30) abandons the await after 30 seconds and returns a structured error response. A module-level asyncio.Semaphore(4) caps concurrent in-flight graph.query calls so that timed-out threads (which continue running in the pool) cannot crowd out other requests by exhausting executor workers. SparqlResponse gains a truncated: bool field so callers can detect a capped result set. (CWE-400)
  • Import upload size limit and extension allowlist (semantica/explorer/routes/export_import.py): No file size or type checks were enforced before reading import uploads. Extension is now validated against {".json", ".csv"} (the formats the handler actually parses — allowlist trimmed to match implementation); a hard 50 MB cap is enforced before content is read. (CWE-434)

  CodeQL / scanning infrastructure

  • Added .github/codeql/codeql-config.yml with paths-ignore for cookbook/**/*.html and cookbook/**/*.js. Notebook-exported HTML files embed entire minified third-party bundles (Plotly + MapLibre GL JS v4.7.1) that triggered false-positive JS alerts #15–#18. The Advanced Setup workflow now references this config via config-file:. Closes CodeQL alerts #15, #16, #17, #18.
  • Removed blanket rule-ID auto-dismiss job from .github/workflows/codeql.yml. The previous job dismissed every open alert whose rule.id matched a fixed list on each main push — this would silently suppress any future real vulnerability of the same type. Replaced with a commented template for pinning specific alert numbers when manual dismissal is genuinely required.

• Fix: Knowledge Explorer — blockers and security hardening (PR #420 by @ZohaibHassan16, review fixes by @KaifAhmad1):

  • Dockerfile: Renamed DockerFile → Dockerfile (case-sensitive filename caused Docker build failures on Linux CI). Fixed CMD module path from the non-existent semantica.server:app to semantica.explorer.app:app, which caused the Docker image to crash on startup. Added app = create_app() at module level in semantica/explorer/app.py so uvicorn can reference the ASGI app instance directly.
  • CORS hardening: Changed EXPLORER_CORS_ORIGINS default from "*" to "http://localhost:5173,http://127.0.0.1:5173". Any deployment that does not explicitly set the env var no longer exposes the API to all origins. The env var override continues to work as before.
  • get_ws_manager() guard (semantica/explorer/dependencies.py): get_ws_manager() now raises HTTP 503 if app.state.ws_manager is absent, matching the existing get_session() guard. Previously raised an unhandled AttributeError during testing or if the lifespan had not completed.
  • SPARQL read-only enforcement (semantica/explorer/routes/sparql.py): Added _is_read_only_query() regex guard — rejects any query whose first keyword is not SELECT, ASK, CONSTRUCT, or DESCRIBE. INSERT, DELETE, UPDATE, LOAD, and DROP queries now return a structured SparqlResponse error instead of being executed against the rdflib projection.
  • Vocabulary import size limit (semantica/explorer/routes/vocabulary.py): Added 10 MB upload cap for both file and raw-text payloads — returns HTTP 413 with a human-readable message before calling parse_skos_file(). Prevents memory exhaustion from oversized RDF uploads.
  • JSON-LD format auto-detection (semantica/explorer/routes/vocabulary.py): Import route now detects .jsonld, .json-ld, and .json file extensions and passes "json-ld" to parse_skos_file(). Previously these extensions fell through to "turtle" and failed silently despite JSON-LD being listed as a supported format.
  • Annotation O(1) lookup (semantica/explorer/routes/annotations.py, semantica/explorer/session.py): create_annotation previously called get_annotations() and scanned the full list to find the just-created annotation (O(N)). Added GraphSession.get_annotation(annotation_id) — O(1) dict lookup — and updated the route to use it directly.
  • Self-loop guard in batchMergeEdges (semantica-explorer/src/store/graphStore.ts): Added if (source === target) continue guard at the start of the loop. The Graphology instance is initialised with allowSelfLoops: false; a self-loop edge from reasoning inferences or provenance cycles previously caused an uncaught Graphology error that silently broke graph loading.
  • Static build artifacts removed from git (.gitignore): Added semantica/static/ to .gitignore and removed all pre-built Vite bundles from version control. The Docker multi-stage build already rebuilds the frontend from source; committing minified bundles bloated repository history and caused merge conflicts on every frontend change.

• Fix: TripletStore.store() IRI resolution regressions (PR #447 follow-up by @KaifAhmad1):

  • Fixed AttributeError crash when entity or relationship IDs are non-string types (e.g. integers emitted by GraphBuilder). _resolve_iri() previously called .startswith() directly on the raw ID; it now coerces any value to str() at entry, restoring the implicit stringification that the old f-string URN minting provided.
  • Fixed W3C vocabulary prefixes (owl:Thing, xsd:date, rdfs:Literal, skos:Concept, etc.) being incorrectly re-namespaced under the ontology base_uri (e.g. https://example.com/owl:Thing) when base_uri was present. _resolve_iri() now consults a known-prefix expansion table (xsd, rdf, rdfs, owl, skos, semantica) before applying base_uri, matching the same prefix map already used in BlazegraphStore. Standard vocabulary IRIs are always expanded to their canonical W3C forms regardless of what base_uri is set to.
  • Added 5 regression tests: integer IDs with and without base_uri, owl:Thing domain/range, xsd:date range, and skos:Concept parent class expansion. Total tests in TestTripletStoreOntologyNamespace: 14.

• Fix: TripletStore.store() ignores ontology namespace base_uri (PR #447 by @KaifAhmad1):

  • store(knowledge_graph, ontology) was minting urn:entity:{id}, urn:class:{type}, and urn:property:{name} URIs for all bare local names, even when ontology.namespace.base_uri was present. This made instance data and ontology class data irreconcilable in SPARQL joins.
  • Extracts base_uri once from ontology["namespace"]["base_uri"] (with ontology["uri"] as fallback) and ensures a trailing separator so concatenation is always a valid IRI path.
  • Introduced _resolve_iri(local, kind) closure applied to all 7 IRI-minting sites: entity URIs, entity types, relationship predicates, ontology class URIs, parent class URIs, property URIs, and domain/range URIs. Explicit entity["uri"] values are never overridden. Falls back to urn: only when no base_uri is available.
  • Added 9 tests in TestTripletStoreOntologyNamespace covering all expansion paths, urn: fallback, explicit URI passthrough, top-level uri key fallback, and trailing-slash safety.

• Fix: Blazegraph literal serialization and SPARQL injection hardening (PR #448 by @KaifAhmad1):

  • Fixed _build_ntriples(), _build_insert_data(), find_triplets(), and delete_triplet() in BlazegraphStore — all four methods previously unconditionally wrapped every triplet object in <...> as an IRI, causing Blazegraph to reject or misparse any triple whose object was a plain string, typed literal, or language-tagged literal.
  • Added _format_object_for_sparql(triplet) — central formatter that selects the correct SPARQL/N-Triples token: IRI (<uri>), typed literal ("value"^^<datatype>), language-tagged literal ("value"@lang), or plain literal ("value").
  • Added _resolve_datatype_iri(datatype) — expands prefixed datatype names (xsd:integer, rdf:langString, rdfs:Literal, owl:real, skos:notation) to their full IRIs instead of producing invalid <xsd:integer> tokens. Accepts full http/https/urn IRIs and already-bracketed IRIs after whitespace validation. Rejects unknown prefixes and bare local names with a clear ValueError.
  • Added language-tag validation against RFC 5646 (^[a-zA-Z]{1,8}(-[a-zA-Z0-9]{1,8})*$) — values containing whitespace, dots, or other punctuation (e.g. "en . CLEAR ALL #") raise ValueError before interpolation, closing a SPARQL injection vector in metadata["lang"] / metadata["language"].
  • Added datatype-string validation — whitespace and SPARQL-delimiting characters inside metadata["datatype"] / metadata["literal_datatype"] raise ValueError, closing the parallel injection vector for typed literals.
  • Added _is_uri_value(value) — URI detection using urlparse; rejects strings that only start with a URI scheme but contain whitespace (e.g. "http not a uri" is serialised as a literal, not an IRI).
  • Added _escape_literal(value) — escapes \, ", \n, \r, \t inside literal strings before SPARQL interpolation.
  • New test file tests/triplet_store/test_blazegraph_store.py — 15 offline unit tests covering URI serialization, plain/typed/language-tagged/escaped literals, prefix expansion, IRI passthrough, injection rejection, and _build_insert_data delegation; all run without a live Blazegraph instance.

• OWLGenerator user-facing schema compatibility fixes (Issue #446):

  • Fixed OWL class/property IRI identifier fallback order to prefer label and then name.
  • Fixed datatype property handling to accept scalar and list range values in rdflib path (including xsd:*, full IRIs, and local names), preventing list-based .startswith() crashes.
  • Fixed generated class/property/domain/range IRIs to use the current ontology dict uri namespace for each generation call (instead of drifting to default namespace manager base URI when per-entity uri is omitted).
  • Fixed subClassOf / subclassOf parent resolution so local class names are expanded to ontology IRIs consistently with domain/range behavior.
  • Added/expanded regression coverage in tests/ontology/test_ontology_comprehensive.py (test_owl_generator_user_facing_schema_compatibility) for label-first fallback, lowercase subclassOf, datatype range lists, and ontology namespace consistency.

• SKOS Vocabulary Module (PR #319 by @KaifAhmad1):

  • Namespace helpers (semantica/ontology/namespace_manager.py): Added get_skos_uri(local_name) — returns the full http://www.w3.org/2004/02/skos/core#<local_name> URI for any SKOS term. Added build_concept_scheme_uri(name) — slugifies a human-readable vocabulary name (spaces/special chars → hyphens, lower-cased) and anchors the result at the configured base URI as <base>/vocab/<slug>.
  • Triplet-store SKOS helpers (semantica/triplet_store/triplet_store.py): Added add_skos_concept(concept_uri, scheme_uri, pref_label, alt_labels, broader, narrower, related, definition, notation) — assembles and stores all required SKOS triples (auto-declares the skos:ConceptScheme, asserts rdf:type skos:Concept, skos:inScheme, skos:prefLabel, and all optional predicates) via the existing add_triplets() API; no new storage paths introduced. Added get_skos_concepts(scheme_uri=None) — issues a SPARQL SELECT via execute_query() and collapses multi-valued altLabel/broader/narrower/related bindings into structured concept dicts; optional scheme_uri restricts results to one vocabulary.
  • OntologyEngine vocabulary APIs (semantica/ontology/engine.py): Added three public methods that delegate to QueryEngine via self.store.execute_query() — list_vocabularies() returns all skos:ConceptScheme instances with labels; list_concepts(scheme_uri) returns every skos:Concept in a scheme with pref_label and alt_labels; search_concepts(query, scheme_uri=None) performs case-insensitive substring matching across skos:prefLabel and skos:altLabel with optional scheme scoping.
  • Security: search_concepts sanitises user input (escapes \, ", newlines) before embedding it in the SPARQL string literal. All URI interpolation uses the existing _sanitize_uri helper.
  • Tests: Added TestSKOSOntologyEngine (14 tests) to tests/ontology/test_ontology_comprehensive.py and TestSKOSTripletStore (6 tests) to tests/triplet_store/test_triplet_store.py. Coverage: URI helpers, vocabulary listing + deduplication, concept listing with multi-value alt-label collapse, search with/without scheme filter, injection sanitisation, empty results, and no-store error paths. 20 new tests, 0 failures, 1162 total passing, 0 regressions.
  • Docs (docs/reference/ontology.md): Added "SKOS Vocabulary Management" section with SKOS data-model reference table, add_skos_concept usage example, bulk import via rdflib + add_triplets, list_vocabularies / list_concepts / search_concepts usage examples, and NamespaceManager URI helper examples.
  • No new top-level Python package created; all code extends existing semantica/ontology/ and semantica/triplet_store/ packages. Fully opt-in and non-breaking.

• SHACL Shape Generation & Validation (PR #318 by @KaifAhmad1):

  • Phase 1 — Generation: Added SHACLGenerator to semantica/ontology/ontology_generator.py — 6-stage internal pipeline: _build_class_index → _generate_node_shapes → _attach_property_shapes → _propagate_inheritance → _apply_quality_tier → serialize. Derives SHACL node and property shapes from any Semantica ontology dict; zero hand-authoring. Three output formats: Turtle, JSON-LD, N-Triples. Three quality tiers: "basic" (structure + cardinality), "standard" (+ sh:in, sh:pattern, inheritance; default), "strict" (+ sh:closed true + sh:ignoredProperties on all non-empty shapes). Iterative inheritance propagation up to 3+ levels, cycle-safe (max 20 passes), no duplicate property shapes per shape. No-domain properties attach to all node shapes. Added PropertyShape, NodeShape, SHACLGraph dataclasses.
  • Phase 1 — Engine API: Added OntologyEngine.to_shacl(ontology, *, format, base_uri, shapes_uri, include_inherited, severity, quality_tier, validate_output) and OntologyEngine.export_shacl(ontology, path, format, encoding) to semantica/ontology/engine.py. Added RDFExporter.export_shacl(shacl_string, file_path, format, encoding) to semantica/export/rdf_exporter.py with extension validation (.ttl, .jsonld, .nt, .shacl).
  • Phase 2 — Runtime Validation: Added SHACLViolation (8 fields: focus_node, result_path, constraint, severity, message, value, shape, explanation; to_dict()) and SHACLValidationReport (conforms, violations, warnings, infos, raw_report; violation_count/warning_count properties; summary(), explain_violations(), to_dict()) to semantica/ontology/ontology_validator.py. Added _run_pyshacl(data_graph_str, shacl_str, data_graph_format, shacl_format) — thin wrapper around pyshacl.validate() returning typed SHACLValidationReport. pyshacl and rdflib are optional deferred imports (pip install semantica[shacl]); ImportError with install hint raised if absent. Added OntologyEngine.validate_graph(data_graph, shacl=None, *, ontology=None, data_graph_format, shacl_format, explain, abort_on_first) — exactly one of shacl/ontology must be provided (ValueError otherwise); explain=True populates plain-English explanations via rule-based templates for all 7 SHACL constraint types (MinCount, MaxCount, Datatype, Class, In, Pattern, Closed).
  • Exports: SHACLGenerator, SHACLGraph, NodeShape, PropertyShape, SHACLValidationReport, SHACLViolation added to semantica/ontology/__init__.py.
  • Security & reliability fixes:
    • High (engine.py): Replaced path-vs-content heuristic (len < 500 and "\n" not in s) with os.path.exists() — prevents attacker-controlled SHACL strings from being silently interpreted as file paths.
    • High (ontology_generator.py): _propagate_inheritance now uses dataclasses.replace(pps) instead of appending parent PropertyShape objects by reference — mutations on a child's inherited property no longer silently affect the parent.
    • Medium (engine.py / ontology_validator.py): Added shacl_format parameter to validate_graph and _run_pyshacl; full format alias map ("ttl"→"turtle", "jsonld"→"json-ld", "ntriples"→"nt") in both to_shacl validate-output and _run_pyshacl — JSON-LD and N-Triples shapes no longer fail parsing.
    • Medium (ontology_generator.py): sh:ignoredProperties now emits full URI <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> instead of prefixed rdf:type — eliminates prefix-dependency in strict-tier Turtle output.
    • Low (ontology_generator.py): _prefix_decls now iterates sorted(graph.prefixes.items()) — deterministic Turtle output for reproducible CI git diff checks.
  • Tests: Added TestSHACLGeneration (16 tests) to tests/ontology/test_ontology_comprehensive.py and TestSHACLHierarchicalAndValidation (18 tests) to tests/ontology/test_ontology_advanced.py. 34 new tests, 0 failures, 1111 total passing, 0 regressions.
  • README: Added ## Unreleased / Coming Next section, SHACL bullet points under Features → Ontology and Export Formats, updated Modules table, full Phase 1 + Phase 2 code examples under ## Ontology, pip install semantica[shacl] under Installation.

• Temporal GraphRAG Integration (PR #402 by @KaifAhmad1):

  • Added TemporalGraphRetriever to semantica/context/context_retriever.py — drop-in wrapper for any ContextRetriever; calls base_retriever.retrieve(query) then filters related_entities/related_relationships via reconstruct_at_time(); at_time=None is a true passthrough; returns new RetrievedContext objects via dataclasses.replace() (no in-place mutation); temporal modules guarded with try/except at import time.
  • Extended ContextRetriever._generate_reasoned_response() and query_with_reasoning() with at_time and header_template parameters — when at_time is set a structured temporal header ([Graph context valid as of: … UTC | Source: KnowledgeGraph snapshot]) is prepended to the LLM context block; omitted when at_time=None (prompt byte-identical to previous behaviour); naive datetimes normalised to UTC; header built via str.replace not .format (format-string injection guard).
  • Added TemporalQueryRewriter and TemporalQueryResult in semantica/kg/temporal_query_rewriter.py — extracts temporal_intent ("before", "after", "at", "during", "between", None), at_time, start_time, end_time, and rewritten_query from natural-language queries; regex-only by default (zero LLM calls), optional LLM-assisted mode; datetime resolution always delegated to TemporalNormalizer; word-boundary guards prevent false matches (at inside that); year fallback handles noun-phrase dates like "the 2021 merger"; never calls reconstruct_at_time.
  • Exported TemporalGraphRetriever from semantica.context; exported TemporalQueryRewriter, TemporalQueryResult from semantica.kg.
  • Security fixes: format-string injection in header template (medium); unconditional temporal module import at package init (low).
  • Bug fixes: in-place mutation of RetrievedContext (high); naive datetime formatted without timezone (low); missing timezone import causing NameError (low).
  • Added 99 tests across tests/context/test_temporal_retriever.py (56) and tests/kg/test_temporal_query_rewriter.py (43); 0 failures, 0 regressions.

• Temporal Provenance & Export (PR #401 by @KaifAhmad1):

  • Transaction time on provenance records (semantica/kg/provenance_tracker.py): track_entity() now automatically attaches recorded_at = datetime.now(UTC).isoformat() to every new record — no opt-in required. Existing records without recorded_at continue to work in all existing query methods (treated as unknown, not an error). Added query_recorded_between(start, end) -> list returning all provenance records whose recorded_at falls within the inclusive range; accepts datetime objects or ISO strings including trailing Z.
  • Fact revision audit trail (semantica/kg/provenance_tracker.py): Added revision_history(fact_id) -> list returning the complete revision chain ordered by recorded_at ascending; each entry includes version (int, 1-based), valid_from, valid_until, recorded_at, author, and optionally revision_type/supersedes; returns [] for unknown facts (never raises). Added export_audit_log(fact_ids, format) -> str supporting "json" (pretty-printed) and "csv" (with header row) formats.
  • OWL-Time RDF export (semantica/export/rdf_exporter.py): export_to_rdf() gains include_temporal: bool = False and time_axis: str = "valid" parameters. When include_temporal=True, emits OWL-Time triples (http://www.w3.org/2006/time#) for every relationship carrying valid_from/valid_until — a time:Interval node linked via time:hasTime, time:hasBeginning/time:hasEnd with time:Instant nodes, and time:inXSDDateTimeStamp values. time_axis controls which axis is exported: "valid", "transaction", or "both". Relationships without temporal metadata are unaffected. Default include_temporal=False produces output identical to current behavior. Design decision for TemporalBound.OPEN: OWL-Time has no standard predicate for "no known end date" — time:hasEnd is omitted and semantica:openEndedInterval "true"^^xsd:boolean is emitted on the interval node instead. Output parses without errors in rdflib.
  • Stable snapshot serialization format (semantica/kg/temporal_query.py, new semantica/kg/schemas/temporal_snapshot_v1.json): create_snapshot() now stamps "format_version": "1.0" on every snapshot. Added validate_snapshot(snapshot) -> bool — validates required fields (format_version, label, timestamp, author, description, entities, relationships, checksum); returns False with structured DEBUG-level error details on failure, never raises. Added migrate_snapshot(snapshot) -> dict — deep-copies and upgrades old-format snapshots to v1.0, populating missing required fields with None; already-v1.0 snapshots returned unchanged with no data loss. New semantica/kg/schemas/temporal_snapshot_v1.json — JSON Schema (draft 2020-12) defining required and optional fields, types, and constraints.
  • Added 28 new tests in tests/test_401_temporal_provenance_export.py covering every acceptance criterion; 451 related tests pass, 0 regressions.

• Temporal Metadata Extraction from Text (PR #400 by @KaifAhmad1):

  • Added extract_temporal_bounds: bool = False parameter to extract_relations_llm(). When True, the LLM prompt is extended with a calibrated confidence scale and four few-shot examples; each returned Relation gains valid_from, valid_until, temporal_confidence (0.0–1.0), and temporal_source_text in its metadata dict. Default False preserves 100% backward compatibility.
  • Confidence scale anchors baked into the prompt: 1.00 = full ISO date, 0.90 = year+month, 0.85 = year only, 0.75 = quarter, 0.65 = named season/approximate range, 0.50 = vague relative with computable anchor, 0.35 = highly vague, 0.00 = no temporal signal. LLMs self-report certainty rather than clustering near 1.0.
  • Low temporal confidence (< 0.5) with a non-null date logs a WARNING; signal is never suppressed — callers decide how to filter.
  • Cache key now includes the extract_temporal_bounds flag to prevent cross-mode cache pollution.
  • Flag propagated through _extract_relations_chunked() so long-text chunked extraction also carries temporal metadata.
  • Added RelationWithTemporalOut and RelationsWithTemporalResponse Pydantic schemas in semantica/semantic_extract/schemas.py. A separate schema is required because RelationOut uses extra="ignore", which silently drops any undeclared field including the four temporal fields.
  • New semantica/kg/temporal_normalizer.py — TemporalNormalizer class (zero LLM calls, pure regex + dateutil arithmetic):
    • normalize(value) → (valid_from, valid_until) UTC datetime tuple or None. Resolution order: ISO 8601 full parse → partial-date regex (year-only, month+year, YYYY-MM, Q[1-4] YYYY) → ambiguous-slash-date detection → domain phrase map → relative phrase resolution via relativedelta.
    • normalize_phrase(phrase) → metadata dict {"maps_to": ..., "type": ..., "domain": [...]} or None — exact match then regex-pattern keys.
    • Ambiguous DD/MM/YYYY-style inputs issue TemporalAmbiguityWarning and return None — never silently guesses locale.
    • Unparseable inputs return None with a debug log — never raise.
    • Relative phrases ("last year", "three months ago", etc.) raise ValueError if reference_date is None rather than guessing.
    • Default phrase map covers 13 domains: General/Policy (effective date, effective from/as of/beginning, in force until, retroactive to, sunset clause), Healthcare (approval date, expiry date, market authorization), Cybersecurity (incident window, campaign period), Supply Chain (certification valid through), Finance (trading halt), Energy (commissioned date, decommissioned date).
    • User-supplied phrase_map is merged over defaults at construction ({**defaults, **user_map}) — custom entries win without forking the library.
  • Added TemporalAmbiguityWarning(UserWarning) to semantica/utils/exceptions.py.
  • Exported TemporalNormalizer from semantica/kg/__init__.py.
  • Added 53 new tests in tests/semantic_extract/test_temporal_extraction.py; zero real LLM calls, suite runs in ~3.5 s. All 873 existing tests continue to pass.

• Fix: OllamaProvider ignores base_url (PR #408 by @AlexeyMyslin, fixed by @KaifAhmad1):

  • OllamaProvider._init_client() was assigning the raw ollama module to self.client instead of instantiating ollama.Client(host=self.base_url), causing all requests to silently hit localhost:11434 regardless of the base_url passed by the user
  • Fixed by replacing self.client = ollama with self.client = ollama.Client(host=self.base_url) — remote Ollama servers (e.g. http://192.168.1.3:11434) are now reachable
  • Added 3 regression tests: default URL forwarded as host, custom URL forwarded as host, and guard ensuring self.client is never the raw module

• Temporal Awareness in Context Graph (PR #399 by @KaifAhmad1):

  • Added valid_from and valid_until fields to the Decision dataclass and record_decision() — decisions now carry explicit validity windows; superseded decisions remain in the graph (history is immutable)
  • Added include_superseded=False and as_of=None parameters to find_precedents_by_scenario() — defaults exclude expired decisions; as_of enables point-in-time precedent queries
  • Added ContextGraph.state_at(timestamp) — returns a serializable point-in-time snapshot of all nodes, edges, and decisions whose validity windows include timestamp; source graph is never mutated
  • Stamped recorded_at on causal relationship edges created via add_causal_relationship() — enables transaction-time filtering
  • Added CausalChainAnalyzer.trace_at_time(event_id, at_time) — reconstructs a causal chain using only edges recorded up to at_time (transaction time); returns an empty list when at_time predates all facts, never raises
  • Added AgentContext.checkpoint(label), diff_checkpoints(label1, label2), and flush_checkpoint(label) — named in-memory context snapshots with structured diffs (decisions_added, decisions_removed, relationships_added, relationships_removed) and optional persistence via TemporalVersionManager
  • Review fixes applied in the same PR:
    • Fixed max_depth error message in trace_at_time to match actual bound (1–100)
    • Fixed Cypher at_time query parameter to RFC3339 UTC (Z suffix) for unambiguous external DB comparisons
    • _normalize_temporal_input now raises ValueError on unparseable strings instead of silently returning raw input
    • Replaced datetime.now() with datetime.utcnow() for all recorded_at and checkpoint timestamps — aligns with codebase convention and avoids wrong local time on Windows
    • flush_checkpoint wraps TemporalVersionManager() construction in a try/except and re-raises as RuntimeError with a clear actionable message
  • Added 7 new tests (93 total across context modules, 0 failures)

• spaCy Runtime Fallback for NER Benchmarks:

  • Hardened NERExtractor spaCy initialization so installed-but-broken spaCy environments no longer crash during extractor construction.
  • Updated ML entity extraction fallback behavior to catch runtime spaCy initialization failures, not just missing-model errors.
  • Added regression coverage for the "spaCy present but unusable at runtime" initialization path.

• Deterministic Temporal Reasoning Engine (PR #398 by @KaifAhmad1, implemented and follow-up fixes by OpenAI Codex):

  • Added semantica.kg.temporal_reasoning as the single source of truth for deterministic, LLM-free temporal reasoning with an explicit zero-LLM module contract
  • Implemented TemporalInterval, full Allen interval algebra via IntervalRelation, and TemporalReasoningEngine
  • Added deterministic helpers for interval overlap/containment checks, open-ended activity checks, interval merging, gap analysis, coverage calculation, timelines, retroactive coverage, and temporal normalization
  • Integrated temporal query interval logic with the reasoning engine in TemporalGraphQuery
  • Preserved semantica.reasoning access via re-exports without making it the canonical implementation source
  • Fixed open-ended query_time_range(..., end_time=None) handling so temporal range queries no longer crash on TemporalBound.OPEN
  • Restored temporal_granularity behavior for point-in-time checks in query_at_time()
  • Eliminated the semantica.reasoning / semantica.kg circular import risk introduced during the initial module move
  • Added regression coverage for all 13 Allen relations, open-ended intervals, month-granularity point queries, open-ended range queries, retroactive coverage, and normalization idempotence

• Temporal Query Engine: Point-in-Time Correctness (PR #397 by @KaifAhmad1, implemented and follow-up fixes by OpenAI Codex):

  • Added reconstruct_at_time(graph, at_time) to TemporalGraphQuery to build a self-consistent point-in-time subgraph without mutating the input graph
  • Updated query_at_time() to use point-in-time reconstruction internally so returned subgraphs exclude dangling edges when entity lifetimes are available
  • Added TemporalConsistencyIssue and TemporalConsistencyReport plus temporal consistency validation for:
    • inverted relationship intervals
    • relationships outside entity lifetimes
    • missing source/target entities
    • overlapping same-type relationships on the same edge
    • temporal gaps where a fact ends and restarts later
  • Added a module-level validate_temporal_consistency(graph) API alongside the query-engine method
  • Implemented sequence and cycle pattern detection with structured outputs containing pattern_type, signature, frequency, and per-occurrence node/edge/time details
  • Implemented calendar-aligned temporal evolution bucketing based on temporal_granularity
  • Added causal ordering controls to find_temporal_paths() via enforce_causal_ordering and ordering_strategy (strict, overlap, loose)
  • Follow-up fixes applied in the same PR:
    • Made validate_temporal_consistency() non-throwing on malformed temporal fields and return report errors instead of raising
    • Enforced exclusive valid_until semantics for point-in-time checks (valid_from <= at_time < valid_until)
    • Kept query_time_range(..., temporal_aggregation="evolution") backward-compatible by returning the flat relationship list plus a new relationship_buckets field
    • Hardened temporal pattern detection for open-ended intervals (TemporalBound.OPEN) to avoid datetime arithmetic/comparison crashes
    • Normalized relationship endpoints during point-in-time reconstruction so mixed-type IDs like 1 and "1" do not silently drop valid edges
    • Added in-code design comments documenting the sequence/cycle output structure required by the checklist
  • Added and expanded regression coverage for point-in-time reconstruction, exclusive end bounds, non-throwing validation, module-level validator access, pattern detection with gap tolerance/open bounds, evolution bucketing, causal ordering, and mixed-type IDs

• Core Temporal Data Model Overhaul (PR #396 by @KaifAhmad1, implemented and follow-up fixes by OpenAI Codex):

  • Added semantica.kg.temporal_model with shared helpers for parsing, normalizing, serializing, and deserializing temporal relationship fields
  • Exported TemporalBound and BiTemporalFact from semantica.kg for backward-compatible temporal relationship handling
  • Updated TemporalGraphQuery to use shared temporal parsing/model helpers instead of ad hoc string handling
  • Added support for valid, transaction, and both time axes in temporal query filtering
  • Standardized temporal normalization on timezone.utc for better cross-version portability
  • Added TemporalValidationError to utils exports and made invalid temporal inputs consistently raise it
  • Added history-preserving temporal revisions in TemporalVersionManager.apply_revision() with provenance metadata and supersession semantics
  • Added safer snapshot persistence by serializing revision metadata before storage and surfacing storage failures as ProcessingError
  • Follow-up fixes applied in the same PR:
    • Added a default factory for BiTemporalFact.recorded_at and preserved legacy transaction-axis behavior by falling back to valid_from when recorded_at is missing
    • Treated TemporalBound.OPEN as an unbounded value in shared query parsing so open-ended facts do not fail in public APIs like analyze_evolution() and path filtering
    • Recomputed snapshot checksums before persisting revised snapshots and any original snapshot inserted during revision flow
    • Replaced second-based revision suffixes with collision-resistant revision IDs/labels to avoid duplicate save failures under rapid revisions
    • Removed warning spam caused by canonical serialized open bounds represented as None
  • Added and expanded regression coverage for UTC normalization, transaction-axis queries, open-ended bounds, revision integrity, checksum verification, and collision-resistant revision identifiers

• Audit Trail, Named Tags, and Rollback Protection (PR #394 by @ZohaibHassan16, reviewed by @KaifAhmad1, follow-up fixes by OpenAI Codex):

  • Added mutation-level audit tracking for ContextGraph node and edge changes via TemporalVersionManager.attach_to_graph() and persistent mutation logging backends
  • Added named version tags in both in-memory and SQLite storage so human-readable tags can point to saved snapshots
  • Added rollback protection to restore_snapshot() so destructive graph restores require explicit confirmation
  • Added get_node_history() for per-entity audit inspection and diff() as a Git-like alias over version comparisons
  • Preserved backward compatibility for snapshot payloads and diff outputs by supporting both nodes/edges and entities/relationships
  • Fixed mixed-schema snapshot comparison and version metadata counts after the audit-trail feature landed on top of PR #393
  • Fixed restore replay so rollback does not generate synthetic mutation events in the audit log
  • Added version-label assignment for previously unlabeled mutations when a snapshot is created
  • Resolved merge conflicts against updated main in managers.py, version_storage.py, context_graph.py, and test_managers.py
  • Added and updated regression coverage for audit history, rollback safety, version-label persistence, and snapshot compatibility

• Snapshot Schema Compatibility Fix (PR #393 by @ZohaibHassan16, reviewed by @KaifAhmad1, follow-up fixes by OpenAI Codex):

  • Fixed silent snapshot restore failures caused by the ContextGraph nodes/edges schema not matching the version manager's legacy entities/relationships expectations
  • Updated temporal snapshot handling to accept both nodes/edges and entities/relationships
  • Preserved both schema shapes in stored snapshots to maintain backward compatibility during migration
  • Fixed temporal diffing and detailed comparison paths so new-format and mixed-format snapshots compare correctly
  • Fixed version metadata counts so entity_count and relationship_count remain accurate for both snapshot schemas
  • Restored ontology snapshot compatibility fields removed during the PR follow-up iteration
  • Added regression coverage for new-format snapshot creation, metadata counts, and mixed-schema diffing

• ContextGraph Traversal Fallbacks for DecisionQuery & DecisionRecorder (PR #386 by @ZohaibHassan16, reviewed and fixed by @KaifAhmad1):

  • Added native ContextGraph fallback execution paths to all 7 DecisionQuery methods (_find_precedents_basic, find_by_category, find_by_entity, find_by_time_range, multi_hop_reasoning, trace_decision_path, find_similar_exceptions) — resolves issue #379 where hardcoded Cypher queries broke in-memory usage
  • Added native ContextGraph fallback paths to 4 DecisionRecorder methods (link_entities, record_exception, link_precedents, _store_decision_node, _store_exception_node) using add_node / add_edge primitives
  • Implemented undirected BFS in multi_hop_reasoning fallback — traverses both outgoing and incoming edges so decisions are reachable from linked entities (matches Cypher (start)-[*1..N]-(d:Decision) semantics)
  • Fixed isinstance(graph_store, ContextGraph) guards → type(graph_store) is ContextGraph — prevents Mock(spec=ContextGraph) from triggering fallback branches and breaking 2 existing tests
  • Fixed add_node(properties=metadata) call in _store_decision_node and _store_exception_node — changed to **metadata so all decision fields are stored flat and remain readable via _dict_to_decision; previous form silently nested every field under a "properties" key
  • Fixed spurious properties={} keyword argument in all add_edge fallback calls — argument did not match the actual add_edge(**properties) signature
  • Fixed tz-aware / naive datetime mismatch in find_by_time_range fallback — strips tzinfo from aware bounds when stored timestamps are naive, preventing TypeError at comparison time
  • Hoisted find_edges() calls out of the BFS while loop in trace_decision_path — edges are now fetched once per call instead of once per visited node, eliminating O(nodes × total_edges) repeated full-graph scans
  • Removed duplicate from ..embeddings import EmbeddingGenerator import in decision_query.py
  • Added tests/context/test_decision_query_fallback.py with 14 tests: full integration test covering the complete fallback flow end-to-end, plus 13 targeted unit tests covering each DecisionQuery and DecisionRecorder fallback method individually, tz-aware/naive datetime mixing, and Mock guard correctness

• ContextGraph Thread Safety & Pagination (PR #385, Issues #378 #376 by @ZohaibHassan16, review & fixes by @KaifAhmad1):

  • ContextGraph: added threading.RLock (self._lock) to __init__; all mutation paths (add_nodes, add_edges, add_node, add_edge, save_to_file, load_from_file, link_graph) and all read/query paths (find_nodes, find_edges, find_node, find_active_nodes, get_neighbors, query, stats, density) now protected with with self._lock: to prevent race-condition corruption under concurrent FastAPI workers
  • find_nodes and find_edges gained native skip/limit pagination parameters so the explorer layer never loads the full collection into memory to slice it
  • GraphSession (session.py): introduced session-level RLock wrapping all graph access; all 8 lazy analytics properties (centrality, community, connectivity, path_finder, node_embedder, similarity, link_predictor, validator) initialised under the lock (thread-safe double-checked); get_nodes() and get_edges() delegate pagination to the graph layer when no in-memory filter is needed
  • pyproject.toml: removed duplicate entry and added missing comma in the all optional-dependency array that caused ERROR Failed to parse pyproject.toml: Unclosed array in CI
  • Fixes applied post-review (by @KaifAhmad1):
    • Fixed /api/graph/search returning empty content and properties — ContextGraph.query() wraps results in node.to_dict() which uses a "properties" envelope, but _node_dict_to_response expected a flat {id, type, content, metadata} shape; session.search() now normalises the envelope before returning
    • Fixed edge metadata silently dropped on import — add_edges() read only from the "properties" key, but edges produced by find_edges() and build_graph_dict() use "metadata"; fixed with edge.get("properties") or edge.get("metadata", {}) fallback
    • Fixed POST /api/enrich/links blocking the asyncio event loop — the O(n) score_link scoring loop ran inline in the async handler; wrapped in asyncio.to_thread(_score_all)
    • Removed merge-artifact dead code in session.py: duplicate self.annotations assignment, duplicate un-locked property set, and double-query logic in get_nodes()/get_edges() that recomputed results outside the lock and threw away the correctly-paginated result computed inside it
    • Removed merge-artifact dead code in enrich.py: unreachable second predict_links implementation block after early return, and duplicate nodes, _ fetch in detect_duplicates

• Knowledge Explorer API Backend (PR #384, Issue #377 by @ZohaibHassan16, review & fixes by @KaifAhmad1):

  • Added semantica.explorer package — a full FastAPI backend for the Semantica Knowledge Explorer dashboard
  • app.py: create_app(session) factory with CORS middleware, custom exception handlers (KeyError→404, ValueError→422), and HTML5 static-file fallback routing; generic Exception handler correctly re-raises HTTPException so dependency-injection 503s are not swallowed
  • session.py: GraphSession — thread-safe container wrapping a ContextGraph with 8 lazily-initialised analytics components (CentralityCalculator, CommunityDetector, ConnectivityAnalyzer, PathFinder, NodeEmbedder, SimilarityCalculator, LinkPredictor, GraphValidator); all lazy properties initialised under RLock to prevent double-instantiation under concurrent requests; shared build_graph_dict(node_ids=None) method eliminates duplication across route files; from_file(path) classmethod loads from JSON
  • ws.py: ConnectionManager — thread-safe WebSocket manager with connect(), disconnect(), broadcast(event_type, data), and send_personal() support; safe disconnection cleanup during broadcast
  • dependencies.py: get_session(request) and get_ws_manager(request) FastAPI Depends-compatible callables; get_session raises HTTP 503 when no session is attached
  • 7 modular route files, all using asyncio.to_thread for sync graph operations:
    • routes/graph.py: GET /api/graph/nodes (type/keyword filter, pagination), GET /api/graph/node/{id}, GET /api/graph/node/{id}/neighbors (BFS, depth 1–5), GET /api/graph/edges (type/source/target filter), GET /api/graph/node/{id}/path (BFS or Dijkstra — algorithm param now correctly dispatched), POST /api/graph/search, GET /api/graph/stats
    • routes/analytics.py: GET /api/analytics (centrality, community, connectivity — comma-separated metrics param), GET /api/analytics/validation
    • routes/decisions.py: GET /api/decisions (category filter, pagination), GET /api/decisions/{id}, GET /api/decisions/{id}/chain (BFS causal chain up to 5 hops), GET /api/decisions/{id}/precedents (category + scenario keyword ranking), GET /api/decisions/{id}/compliance (in-graph check over violates/non_compliant/breaches edges — no longer a stub)
    • routes/temporal.py: GET /api/temporal/snapshot (ISO-8601 at param), GET /api/temporal/diff (added/removed node sets between two timestamps), GET /api/temporal/patterns (graceful fallback when TemporalPatternDetector unavailable, with warning log for unexpected errors)
    • routes/enrich.py: POST /api/enrich/extract (NLP entity/relation extraction), POST /api/enrich/links (per-node link prediction via score_link against all non-adjacent candidates — fixed from broken predict_links call), POST /api/enrich/dedup (duplicate detection — fixed missing asyncio.to_thread that was blocking the event loop), POST /api/reason (forward/backward inference via Reasoner)
    • routes/export_import.py: POST /api/export (12 formats: JSON, Turtle, RDF-XML, N-Triples, CSV, GraphML, GEXF, OWL, Cypher, AQL, YAML — temp file always cleaned up via try/finally), POST /api/import (JSON/JSON-LD multipart upload with WebSocket progress events)
    • routes/annotations.py: GET /api/annotations, POST /api/annotations (validates node exists; add_annotation mutates dict in-place so no extra roundtrip), DELETE /api/annotations/{id}
  • schemas.py: 28 Pydantic v2 request/response models covering all endpoint shapes including pagination, temporal, enrichment, compliance, and annotation types
  • __init__.py: semantica-explorer CLI entry point — --graph, --host, --port, --no-browser args; validates graph file exists; checks for uvicorn; opens browser after 1.5 s delay
  • pyproject.toml: added [project.optional-dependencies] explorer group (fastapi, uvicorn[standard], websockets, python-multipart); registered semantica-explorer script entry point; fixed missing comma in all extra that broke pip install semantica[all]
  • Fixes applied post-review (by @KaifAhmad1):
    • Fixed predict_links endpoint — was calling predictor.predict_links(graph_dict, node_id, top_n=...) with wrong type (dict as graph_store), wrong positional arg (node_id as node_labels), and wrong kwarg (top_n vs top_k); rewrote to iterate all non-adjacent candidate nodes and call predictor.score_link(session.graph, source, candidate) directly
    • Fixed detect_duplicates endpoint — session.get_nodes() was called directly in an async def handler without asyncio.to_thread, blocking the event loop
    • Fixed temp file leak in export_graph — file was not deleted on exception from export_fn or open(); wrapped in try/finally; moved import os to module level
    • Fixed pyproject.toml all extra — two consecutive strings with no comma between them caused a TOML syntax error
    • Fixed generic Exception handler swallowing HTTPException(503) raised by get_session
    • Fixed compliance endpoint — imported PolicyEngine then discarded it, always returning compliant=True; replaced with in-graph edge scan
    • Fixed temporal_patterns bare except Exception silently hiding bugs — split into ImportError (silent graceful) and Exception (warning log)
    • Fixed all 8 lazy analytics properties to initialise under _lock (thread-safe double-checked)
    • Fixed find_path ignoring the algorithm query param — now dispatches to dijkstra_shortest_path or bfs_shortest_path
    • Removed unnecessary get_annotations() round-trip in create_annotation
    • Removed import traceback unused import in app.py
    • Deduplicated _build_graph_dict (was copied identically in graph.py, analytics.py, export_import.py) into GraphSession.build_graph_dict()
  • 49 integration tests in tests/explorer/test_explorer_api.py using starlette.testclient.TestClient — all passing; covers health, nodes, edges, search, stats, decisions, causal chains, precedents, compliance (including violation detection), temporal snapshots/diff/patterns, analytics, reasoning, entity extraction, link prediction, deduplication, annotations, export (JSON + node-subset), and import (JSON + edges + unsupported format)

• Reasoning Dead Code Removal (PR #387, Issue #382 by @ZohaibHassan16):

  • Removed lines 357–358 in semantica/reasoning/reasoner.py that silently overwrote the sophisticated _match_pattern regex (which handles pre-bound variable embedding, repeated-variable backreferences via (?P=var), and non-greedy named capture groups) with a simpler re.escape-based pattern, making all the prior logic unreachable dead code
  • Removed duplicate unreachable return None on line 368 (syntactically dead, appearing immediately after another return None in the same branch)
  • Surfaced re.error exceptions instead of swallowing them with except Exception: pass, preventing silent failures when malformed patterns were passed to re.match
  • Before this fix, any rule using the same variable twice (e.g. rel(?x, ?x)) generated a duplicate named group error that was silently caught, causing the match to return None regardless of the fact — breaking transitivity, symmetry, and self-join rule patterns entirely

• Agno Agentic Framework Integration (Issue #249):

  • Added AgnoContextStore — graph-backed agent memory implementing the agno.memory.db.base.MemoryDb protocol; wraps AgentContext + VectorStore; supports create(), table_exists(), memory_exists(), read_memories(), upsert_memory(), delete_memory(), drop_table(), clear() plus extended record_decision(), find_precedents(), retrieve() methods
  • Added AgnoKnowledgeGraph — multi-hop GraphRAG knowledge base implementing agno.knowledge.base.AgentKnowledge; ingests files, directories, URLs, and raw text via NER → relation extraction → graph build → vector index pipeline; search() returns AgnoDocument objects; get_graph_context(entity) returns text summary of entity's graph neighbourhood
  • Added AgnoDecisionKit — Agno Toolkit subclass exposing 6 decision-intelligence tools: record_decision, find_precedents, trace_causal_chain, analyze_impact, check_policy, get_decision_summary
  • Added AgnoKGToolkit — Agno Toolkit subclass exposing 7 KG pipeline tools: extract_entities, extract_relations, add_to_graph, query_graph, find_related, infer_facts, export_subgraph
  • Added AgnoSharedContext — team-level coordinator with a single shared ContextGraph; bind_agent(role) returns a role-scoped _AgentScopedStore with cross-agent memory visibility; thread-safe via RLock
  • All 5 components degrade gracefully when agno is not installed (AGNO_AVAILABLE flag); importable and functional without agno present
  • Added agno = ["agno>=1.0.0"] optional dependency in pyproject.toml; included in all extra
  • 110 integration tests in tests/integrations/agno/ covering all public APIs, MemoryDb protocol compliance, GraphRAG search, tool registration, shared memory isolation, and thread-safety
  • 3 cookbook notebooks in cookbook/integrations/: agno_decision_intelligence.ipynb (loan underwriting), agno_graphrag_context.ipynb (regulatory compliance), agno_multi_agent_shared_context.ipynb (multi-agent team coordination)
  • Full reference documentation in docs/integrations/agno.md

• Novita AI Provider (PR #374 by @Alex-wuhu):

  • Added NovitaProvider — OpenAI-compatible integration via https://api.novita.ai/v1; supports generate() and generate_structured() (JSON forced format)
  • Default model: deepseek/deepseek-v3.2; configurable via NOVITA_API_KEY environment variable
  • Registered "novita" in the built-in provider factory; usable via create_provider("novita")
  • Added integration tests in tests/test_novita_integration.py with proper assertions and graceful skip when NOVITA_API_KEY is unset

• Native Datalog Reasoning Engine (PR #371, Issue #368 by @ZohaibHassan16, reviewed and fixed by @KaifAhmad1):

  • Added DatalogReasoner to semantica.reasoning — a pure-Python, bottom-up semi-naive fixpoint engine with guaranteed termination on finite graphs
  • Supports recursive Horn clause rules (e.g. ancestor(X,Y) :- parent(X,Z), ancestor(Z,Y).) that existing engines loop on indefinitely
  • Memory-optimized _unify() with deferred dict allocation — zero allocation on failed unifications
  • O(1) delta-index lookup per iteration eliminates redundant O(N) rule re-evaluations in semi-naive loop
  • query("pred(?X, ?Y)") returns variable-binding dicts; supports both uppercase ?Y and lowercase ?y variable syntax
  • query(..., bindings={"Y": "val"}) pre-binds variables for exact-match verification
  • load_from_graph(ContextGraph) converts all edges and nodes to Datalog facts in one call; handles both find_edges/find_nodes and raw edges/nodes graph APIs
  • add_fact() accepts "pred(a, b)" strings and Semantica dicts (subject/predicate/object, source/target/type, type/id shapes); warns on unrecognised dict format instead of silently dropping
  • _derived cache flag — derive_all() skips re-evaluation when no facts or rules have changed since last run; query() respects the cache
  • Progress tracking wrapped in try/finally — stop_tracking() always called even on exception
  • DatalogReasoner, DatalogFact, DatalogRule exported from semantica.reasoning
  • 18 tests covering recursive rules, multi-hop inference, variable binding, graph integration, idempotency, and edge cases — all passing

• Ontology Diff & Migration (PR #367 by @ZohaibHassan16, review & fixes by @KaifAhmad1):

  • VersionManager.diff_ontologies(base, target) — structured diff between two ontology dicts using hash-map lookups; handles URI-less items via name fallback; deep equality checks for unordered lists; now covers classes, properties, individuals, and axioms
  • ChangeLogAnalyzer.analyze(diff) — classifies each change by semantic impact: removed classes/properties → CRITICAL/BREAKING; narrowed domain/range/cardinality → HIGH/BREAKING; hierarchy modifications → MEDIUM/POTENTIALLY_BREAKING; added elements and annotation updates → INFO/NON_BREAKING
  • ImpactReport dataclass and generate_change_report(diff) public helper — returns a structured dict with summary, impact_classification (breaking / potentially_breaking / safe), recommendations, and the raw diff
  • OntologyEngine.compare_versions(base_id, target_id, **options) — end-to-end orchestrator: loads versions from VersionManager, runs diff_ontologies, generates impact report; accepts base_dict/target_dict overrides to bypass version store; run_validation=True triggers OntologyValidator on the target schema; graph_data=... additionally runs GraphValidator on instance data against the new schema
  • OntologyEngine.get_ontology_version_dict(version_id) — utility to load a registered version as a plain dict ready for diffing
  • Documentation added to docs/reference/change_management.md: "Ontology Diff & Migration" section with code example and full report format reference
  • 7 tests added to tests/change_management/test_managers.py covering: empty diff, unordered list equality, URI/name fallback, breaking class removal, narrowed domain (HIGH), safe additions and annotation changes, compare_versions dict override, version-not-found error path, individuals/axioms diff coverage, null constraint value flagged as breaking
  • Fixes applied post-review (by @KaifAhmad1):
    • Fixed typo in ChangeCategory enum value: "potenitally_breaking" → "potentially_breaking"
    • Fixed missing space in impact description string: f"New{entity_type}" → f"New {entity_type}"
    • Added null-value guard in _analyze_field_changes — constraint fields with None old/new value are now correctly flagged as breaking instead of silently passing the subset check
    • Made ChangeLogAnalyzer stateless — report is now a local variable passed into _generate_recommendations(report) rather than stored as self.report; removes re-entrancy hazard
    • Removed no-op __init__ from ChangeLogAnalyzer
    • Replaced non-portable emoji markers in recommendations (✘✘✘, ¤¤¤, ☺☺☺) with plain-text tags ([BREAKING], [WARNING], [SAFE])
    • Extended diff_ontologies to cover individuals and axioms — previously only classes and properties were diffed; the public compare_versions path now returns all four element types
    • Fixed exception chaining in compare_versions: raise ProcessingError(...) from e to preserve original traceback
    • Removed silent ImportError swallow for GraphValidator — it is a first-party module; an ImportError indicates a broken install, not a graceful skip
    • Added comment on deferred VersionManager import in OntologyEngine.__init__ explaining the circular-import constraint
    • Fixed import-before-docstring in tests/change_management/test_managers.py
    • Fixed broken Markdown link syntax in docs JSON example block: "[http://...](http://...)" → bare URI string
    • Updated docs recommendations example to match the new plain-text tag format

• Ontology Alignment API (PR #361 by @ZohaibHassan16, review & fixes by @KaifAhmad1):

  • Alignment representation using standard RDF predicates: owl:equivalentClass, owl:equivalentProperty, owl:sameAs, skos:exactMatch, skos:closeMatch, skos:broadMatch, skos:narrowMatch, skos:relatedMatch
  • OntologyEngine.create_alignment(source_uri, target_uri, predicate) — store alignment triples in TripletStore
  • OntologyEngine.get_alignments(entity_uri) — bidirectional retrieval of all alignments for an entity
  • OntologyEngine.list_alignments(ontology_uri=None) — list all alignments, optionally filtered by ontology namespace
  • NamespaceManager.get_alignment_predicates() — expose standard OWL/SKOS alignment URIs as a convenience dict
  • ReuseManager.suggest_alignments(target, source) — O(N+M) hashmap heuristic to suggest alignments based on exact label matches across ontologies
  • ReuseManager.merge_ontology_data(..., compute_alignments=True) — optionally attach suggested alignments to merge output without auto-committing unverified triples
  • QueryEngine.expand_entity_uri(uri, store, use_alignments=True) — bidirectional SPARQL expansion to include aligned equivalents; no-ops when flag is False
  • QueryEngine.build_values_clause(variable, uris) — generate a SPARQL VALUES clause for injecting expanded URIs into queries
  • Alignment-aware queries section added to docs/reference/triplet_store.md
  • Ontology Alignment section added to docs/reference/ontology.md
  • Fixes applied post-review (by @KaifAhmad1):
    • Fixed progress tracker leak in expand_entity_uri — stop_tracking was only called inside the hasattr(execute_sparql) branch; backends without it silently leaked a tracker entry
    • Fixed relatedMatch predicate gap — get_alignment_predicates() exposed skos:relatedMatch but all three SPARQL FILTER lists omitted it, making those alignments permanently invisible
    • Fixed SPARQL injection in list_alignments — previously only " was escaped; \, {, and } are now also percent-encoded to prevent WHERE block breakout
    • Fixed SPARQL injection in build_values_clause — URIs now run through _sanitize_uri before wrapping in angle-bracket literals
    • Added full-URI validation in create_alignment — raises ProcessingError if predicate is a CURIE instead of a full URI, preventing silent storage of unqueryable triples
    • Fixed E2E test test_end_to_end_cross_ontology_uri_flow — previously mocked the method under test; now uses a real mock backend with execute_sparql to exercise the actual expansion and VALUES clause injection flow
  • 19 tests added covering: create_alignment, get_alignments, suggest_alignments, merge with alignment computation, expand_entity_uri (enabled/disabled), build_values_clause, and full E2E cross-ontology query flow

• Context Explainability Output Fixes (by @KaifAhmad1):

  • Fixed decision-node storage in ContextGraph so full human-readable scenario, reasoning, and decision metadata are preserved on graph nodes instead of degrading into opaque IDs or truncated display text
  • Fixed causal and precedent reconstruction paths in the context module so returned Decision objects prefer readable stored fields over raw node identifiers
  • Fixed context aggregate outputs to return enriched readable payloads for influence, causality, similarity, policy-impact, and entity-similarity workflows instead of bare UUID lists or tuple-only results
  • Fixed PolicyEngine.get_affected_decisions() so both Cypher and fallback branches return consistent decision metadata including scenario, category, outcome, and confidence
  • Fixed EntityLinker similarity flows so enriched similarity results are consumed correctly across internal linking paths and public search aliases
  • Fixed CentralityCalculator._build_adjacency() to handle ContextGraph edges (dataclass ContextEdge objects with source_id/target_id) so calculate_degree_centrality() and related centrality algorithms work correctly when a ContextGraph is passed as the graph store
  • Fixed downstream KG integrations in node_embeddings, link_predictor, centrality_calculator, path_finder, and context retrieval fallbacks to normalize enriched neighbor/node outputs without breaking graph algorithms
  • Added 23 regression tests in tests/context/test_context_explainability_regression.py covering readable decision text preservation, enriched causal/path outputs, policy-impact results, entity similarity payloads, and compatibility with KG consumers

[0.3.0] - 2026-03-10

• Context Graph Feature Completeness (by @KaifAhmad1):

  • Added valid_from / valid_until temporal validity fields to ContextNode and ContextEdge dataclasses — both expose is_active(at_time=None) -> bool; nodes/edges without these fields are always considered active
  • Added add_node(valid_from=..., valid_until=...) and add_edge(valid_from=..., valid_until=...) support — validity windows are extracted from **properties and stored as first-class dataclass fields, not in metadata
  • Added ContextGraph.find_active_nodes(node_type=None, at_time=None) — returns only nodes whose validity window includes the given time (defaults to datetime.utcnow()); complements find_nodes() with temporal filtering
  • Added min_weight: float = 0.0 parameter to ContextGraph.get_neighbors() — edges with weight below the threshold are skipped during BFS traversal, enabling weighted/confidence-filtered multi-hop navigation; fully backward-compatible (default 0.0 passes all edges)
  • Added ContextGraph.link_graph(other_graph, source_node_id, target_node_id, link_type="CROSS_GRAPH") -> str — creates a navigable bridge between two separate ContextGraph instances; records a marker edge internally and returns a link_id
  • Added ContextGraph.navigate_to(link_id) -> (other_graph, target_node_id) — resolves a link_id to the target graph and its entry node, enabling hierarchical cross-graph traversal (e.g. agent moving from a high-level decision graph into a domain-specific sub-graph)
  • Added ContextGraph.resolve_links(registry) — reconnects cross-graph links after load_from_file(); save_to_file() now persists a links section with other_graph_id so navigation survives the full save/load cycle
  • Added graph_id field to ContextGraph — stable UUID per instance, persisted to JSON, so separate graphs can identify each other after reload
  • Fixed is_active() on ContextNode and ContextEdge — tz-aware datetime inputs are now normalised to tz-naive UTC before comparison, preventing TypeError when callers pass datetime.now(timezone.utc)
  • Fixed valid_from / valid_until serialisation — add_nodes(), add_edges(), to_dict(), and from_dict() all now preserve and restore validity windows; previously these fields were silently lost
  • Fixed cross-graph link artifact — link_graph() now pre-creates a "cross_graph_link" typed ContextNode for the marker before inserting the marker edge, preventing _add_internal_edge() from auto-creating a phantom "entity" node
  • Added 14 tests in tests/context/test_cross_graph_navigation.py covering link creation, phantom-node prevention, and full save/load round-trips with resolve_links()
  • Fixed pipeline_builder.add_step() return type annotation from "PipelineBuilder" to "PipelineStep" — implementation was already correct per 0.3.0-beta changelog, only signature and docstring were stale
  • Fixed test_hybrid_search_performance timing computation — accumulated a real search_times list and compute true average; raised threshold to < 5.0s to account for real sentence-transformers (384-dim) latency

• 0.3.0 Bug Fixes & Comprehensive Real-World Tests (by @KaifAhmad1):

  • Fixed ProvenanceTracker missing from semantica/kg/__init__.py exports — from semantica.kg import ProvenanceTracker now works correctly
  • Fixed duplicate relation creation in _parse_relation_result — orphaned legacy block was appending every relation twice; removed the duplicate block
  • Added extraction_method parameter to _parse_relation_result; typed extraction path now correctly sets "llm_typed" instead of "llm" in relation metadata
  • Fixed cross-test cache pollution in tests/semantic_extract/test_retry_logic.py — module-level _result_cache now cleared in setUp() to prevent intermittent failures when tests share input text
  • Added tests/test_030_realworld_comprehensive.py: 85 real-world tests covering all 0.3.0-alpha/beta features with real data (tech companies, CEOs, products, investment chains, healthcare scenarios)
    • ContextGraph basic operations and decision tracking lifecycle
    • KG algorithms: centrality, community detection, embeddings, path finding, similarity, link prediction, connectivity
    • PolicyEngine, DecisionQuery, AgentContext, Decision model serialization
    • ProvenanceTracker with GraphBuilderWithProvenance and AlgorithmTrackerWithProvenance
    • Deduplication v2 with blocking strategies, RDF/TTL export, Reasoner inference
    • Pipeline builder/validator/failure handler with retry policies
    • Multi-hop investment chain (Microsoft→OpenAI, Google→Anthropic) end-to-end
    • Healthcare entity extraction and knowledge graph construction E2E

[0.3.0-beta] - 2026-03-07

• Multi-Founder LLM Extraction & Reasoner Inference Fix (PR #354 by @KaifAhmad1):

  • Fixed _parse_relation_result in methods.py — unmatched subjects/objects now produce a synthetic UNKNOWN entity instead of silently dropping the relation; all LLM-returned co-founders are preserved
  • Rewrote _match_pattern in reasoner.py — splits pattern on ?var placeholders first, then escapes only the literal segments; pre-bound variables resolve to exact literals, repeated variables use backreferences, non-greedy .+? prevents over-consumption of literal separators
  • Added tests/reasoning/test_reasoner.py with 4 tests covering multi-word value inference, pre-bound variables, binding conflicts, and single-word regression
  • Added tests/semantic_extract/test_relation_extractor.py with 6 tests covering all-founders returned, synthetic entity creation, matched entity integrity, predicate/confidence preservation, empty response, and malformed entries

• TTL Export Alias Fix (PR #355 by @KaifAhmad1):

  • Added _format_aliases map in RDFExporter so format="ttl", "nt", "xml", "rdf", and "json-ld" resolve to their canonical counterparts without breaking existing callers
  • Alias resolution applied at the top of export_to_rdf() before format validation — zero public API changes
  • Added working TTL export cell to cookbook/introduction/15_Export.ipynb (Step 3: RDF Export)
  • Added tests/export/test_rdf_exporter.py with 8 tests covering all aliases, canonical formats, error handling, and file export

• Incremental/Delta Processing Feature (PR #349 by @ZohaibHassan16, reviewed and fixed by @KaifAhmad1):

  • Native delta computation between graph snapshots using SPARQL queries
  • Delta-aware pipeline execution with delta_mode configuration for processing only changed data
  • Version snapshot management with graph URI tracking and metadata storage
  • Snapshot retention policies with automatic cleanup via prune_versions() method
  • Integration with pipeline execution engine for incremental workflows
  • Significant performance improvements: processes only changes instead of full datasets
  • Cost optimization: dramatically reduces compute and storage requirements for large-scale operations
  • Production-ready for near real-time pipelines and frequent deployment scenarios
  • Bug fixes: corrected SPARQL variable order, fixed class references, resolved duplicate dictionary keys
  • Comprehensive test coverage including delta mode integration tests
  • Complete documentation with usage examples and API references
  • Essential for enterprise-grade, large-scale semantic infrastructure

• Deduplication v2 Migration Guide (PR #344 by @ZohaibHassan16, fixes by @KaifAhmad1):

  • Added comprehensive MIGRATION_V2.md documentation for Deduplication v2 Epic #333
  • Documented Candidate Generation V2 with multi-key blocking and phonetic matching
  • Documented Two-Stage Scoring prefilter with configurable thresholds
  • Documented Semantic Relationship Deduplication v2 with synonym mapping
  • Added practical code examples for all V2 features with opt-in configuration
  • Fixed critical infinite recursion bug in dedup_triplets() function
  • Completed Epic #333 with comprehensive migration path and documentation
  • Performance: 5.86x speedup confirmed (129ms vs 754ms) for semantic deduplication
  • Full backward compatibility maintained with legacy mode as default

• Semantic Relationship Deduplication v2 (PR #340 by @ZohaibHassan16, fixes by @KaifAhmad1):

  • Implemented opt-in semantic relationship deduplication mode (semantic_v2) with 6.98x performance improvement
  • Added canonicalization engine with predicate synonym mapping (works_for → employed_by)
  • Implemented fast-path O(1) hash matching for exact canonical signature comparisons
  • Added weighted semantic scoring (60% predicate + 40% object composition) with explainable semantic_match_score metadata
  • Enhanced dedup_triplets() function as first-class API in methods.py
  • Integrated semantic deduplication into merge strategy with canonical key generation
  • Added literal normalization for whitespace cleanup in object matching
  • Maintained full backward compatibility with legacy mode as default
  • Fixed critical infinite recursion bug in dedup_triplets() function via registry name checking
  • Performance: Semantic V2 (~83ms) vs Legacy (~579ms) - 6.98x speedup confirmed
  • All 13 deduplication benchmarks passing with comprehensive test coverage

• Two-Stage Scoring Prefilter (PR #339 by @ZohaibHassan16):

  • Implemented opt-in two-stage scoring with fast prefilter gates to eliminate expensive semantic scoring for obvious non-matches
  • Prefilter gates: type mismatch detection, name length ratio validation, token overlap requirements
  • Performance improvements: 18-25% faster batch processing with prefilter enabled
  • Configurable thresholds: min_length_ratio, min_token_overlap_ratio, required_shared_token
  • Enhanced explainability with score breakdown and rejection reasons in metadata
  • Complete backward compatibility with default prefilter_enabled=False

• Candidate Generation v2 with Multi-Key Blocking (PR #338 by @ZohaibHassan16):

  • Implemented opt-in candidate generation strategies (legacy, blocking_v2, hybrid_v2) to address O(N²) pair explosion during deduplication
  • Multi-key blocking with normalized token prefixes, type-aware keys, and optional phonetic (Soundex) blocking
  • Deterministic candidate budgeting with max_candidates_per_entity limit using stable sorting
  • Efficient pair generation with set-based deduplication across overlapping blocks
  • Performance improvements: 63.6% faster in worst-case scenarios (0.259s → 0.094s for 100 entities)
  • Complete backward compatibility with default candidate_strategy="legacy"
  • Added configuration options: blocking_keys, enable_phonetic_blocking, max_candidates_per_entity

• ArangoDB AQL Export Support (PR #342 by @tibisabau):

Added

• ArangoDB AQL Export Support (PR #342 by @tibisabau)

  • Full-featured ArangoDB AQL exporter with 642 lines of production-ready code
  • Comprehensive AQL INSERT statement generation for vertices and edges
  • Configurable collection names with validation and sanitization
  • Batch processing support for large knowledge graphs (default: 1000)
  • Added export_arango() convenience function for easy access
  • Enhanced unified export with AQL format support and .aql auto-detection
  • Added export_arango() convenience function for easy access
  • Enhanced unified export with AQL format support and .aql auto-detection
  • Integrated with method registry for extensibility
  • 17 comprehensive test cases with 100% pass rate
  • Enterprise-grade ArangoDB multi-model database integration

• Apache Parquet Export Support (PR #343 by @tibisabau):

• Apache Parquet Export Support (PR #343 by @tibisabau)

  • Full-featured Apache Parquet exporter with 701 lines of production-ready code
  • Columnar storage format optimized for analytics and data warehousing
  • Configurable compression codecs (snappy, gzip, brotli, zstd, lz4, none)
  • Explicit Arrow schemas with type safety and consistency
  • Field normalization for varied entity and relationship naming conventions
  • Structured metadata handling using Parquet struct fields
  • Added export_parquet() convenience function for easy access
  • Enhanced unified export with Parquet format support and .parquet auto-detection
  • Added export_parquet() convenience function for easy access
  • Enhanced unified export with Parquet format support and .parquet auto-detection
  • Integrated with method registry for extensibility
  • 25 comprehensive test cases with 100% pass rate
  • Enterprise-grade analytics integration with pandas, Spark, Snowflake, BigQuery, Databricks

Fixed

• Fixed NameError: missing Type import in utils/helpers.py

• Fixed NameError: missing Type import in utils/helpers.py

  • Added Type to typing imports to fix retry_on_error decorator
  • Removed unused Type import from config_manager.py
  • Resolves ImportError when importing semantica modules
  • Fixes capability gap analysis notebook execution

• Test Suite Fixes: 0.3.0-alpha & Unreleased Features (PR utils by @KaifAhmad1):

  Context Module (semantica/context/)

  • Fixed retrieve_decision_precedents to gate entity extraction on use_hybrid_search=True — was incorrectly extracting entities when flag was False
  • Fixed _extract_entities_from_query to use word[0].isupper() instead of word.istitle() — correctly captures CreditCard, CustomerID etc.
  • Added missing expand_context method — BFS graph traversal via knowledge_graph.get_neighbors
  • Added missing _get_decision_query method — creates a DecisionQuery from the knowledge graph
  • Fixed hybrid_retrieval to call expand_context(query) once (not per-entity) and include "query" key in return dict
  • Fixed dynamic_context_traversal to call expand_context once per query instead of per entity
  • Fixed multi_hop_context_assembly to use _get_decision_query() for robust decision lookup
  • Fixed _retrieve_from_vector to fall back to result["metadata"]["content"] when result["content"] is absent — prevents empty content and negative similarity scores during semantic re-ranking

  Knowledge Graph Module (semantica/kg/)

  • Fixed calculate_pagerank — added alpha and max_iter parameter aliases; changed return format to structured dict {"centrality": scores, "rankings": sorted_list}
  • Fixed community_detector._to_networkx to return a NetworkX graph directly when one is passed (was converting to adjacency list, silently losing all edges)
  • Added method as alias for algorithm parameter in detect_communities
  • Fixed _build_adjacency to handle "edges" key (list of tuples) in addition to "relationships" (list of dicts)
  • Added _track_generic base method and 9 domain-specific tracking methods to AlgorithmTrackerWithProvenance: track_influence_analysis, track_verification_analysis, track_supply_chain_paths, track_bottleneck_analysis, track_quality_analysis, track_lead_time_analysis, track_cross_domain_analysis, track_cross_domain_similarity, track_collaboration_potential
  • Created new provenance_tracker.py module with ProvenanceTracker class (track_entity, get_all_sources, clear)

  Pipeline Module (semantica/pipeline/)

  • Fixed execution_engine retry loop to properly iterate up to max_retries (was only retrying once regardless of policy)
  • Added RecoveryAction dataclass and handle_failure(error, policy, retry_count) method to FailureHandler — implements LINEAR, EXPONENTIAL, and FIXED backoff strategies
  • Fixed pipeline_builder.add_step to return the created PipelineStep object instead of self
  • Added validate as a public alias for validate_pipeline in PipelineValidator
  • Updated missing-dependency error message to "Missing dependency '{dep}' for step '{name}'" for consistent test assertions

  Vector Store (semantica/vector_store/)

  • Relaxed test_batch_processing_performance threshold from < 100ms to < 500ms per decision — original threshold was too tight for development machines running a real sentence-transformers embedding model (384-dim)

  Test File Fixes

  • test_end_to_end_context_integration.py — replaced emoji characters (✅, ❌, 🔄, ⚠️) with ASCII equivalents ([OK], [FAIL], [...], [WARN]) to fix Windows cp1252 encoding error
  • test_context_retriever_precedents.py — moved assert_called_once_with inside with patch.object block; fixed assertion to use decision.scenario not decision.decision_id; removed "iPhone" (lowercase-first) from entity extraction assertion
  • test_real_world_scenarios.py — fixed duplicate source= keyword argument (renamed to label=); fixed cross-domain analysis loop to iterate over all social network users instead of only academic_users
  • test_pipeline_comprehensive.py — changed test_pipeline_validator_missing_deps to call validator.validate(builder) directly instead of builder.build() which raises ValidationError before validation can complete

  Results: ~840 tests passing, 36 skipped (external services), 0 failed

[0.3.0-alpha] - 2026-02-19

Added / Changed

• Decision Tracking System: Complete decision lifecycle management with audit trails and provenance tracking
• Advanced KG Algorithms: Node2Vec embeddings, centrality analysis, community detection for decision insights
• Enhanced Context Module: Unified AgentContext with granular feature flags and decision tracking integration
• Vector Store Features: Hybrid search combining semantic, structural, and category similarity
• Policy Management: Versioning, compliance checking, and exception handling
• Production Ready Architecture: Scalable design with comprehensive error handling and validation

Fixed

• Fixed import issues in test suite (ProvenanceTracker location fixes)
• Fixed causal analyzer validation (max_depth bounds checking)
• Fixed test compatibility with updated method signatures
• Fixed mock object setup in test suites
• Comprehensive test suite fixes for decision tracking features

Testing

• 113+ tests passing across context and core modules
• Comprehensive decision tracking test coverage
• Enhanced error handling and edge case testing
• Fixed all critical test failures for release readiness

Documentation

• Enhanced context module documentation

• Updated API references for decision tracking features

• Comprehensive usage guides and examples

• Fixed: Context Graphs decision tracking bugs and added comprehensive test coverage (PR #315 by @KaifAhmad1)

  • Fixed empty/None decision ID handling in ContextGraph.add_decision()
  • Fixed None metadata handling to prevent TypeError
  • Fixed causal chain depth logic and node exclusion
  • Fixed nonexistent node handling in add_causal_relationship()
  • Added missing properties field in to_dict serialization
  • Added missing from_dict method for graph deserialization
  • Fixed precedent search direction in find_precedents()
  • Fixed UUID generation logic in all decision models
  • Added comprehensive test suite with 9 tests covering all features
  • All 71 context tests now passing (100% success rate)

• Fixed: PolicyEngine latest version selection on ContextGraph; AgentContext fallback robustness and secure logging (PR #TBD by @KaifAhmad1)

• Tests: Added ContextGraph fallback and AgentContext smoke tests; full suite passing

  • Apache AGE Backend Security Fixes (PR #311 by @Sameer6305, fixes by @KaifAhmad1):
  • Added AgeStore class with GraphStore API compatibility
  • Fixed SQL injection vulnerabilities with input validation
  • Added psycopg2-binary dependency and migration guide
  • Fixed parameter replacement and test mock leakage
  • Enhanced error handling and Unicode display issues

• Context Engineering Enhancement (PR #307 by @KaifAhmad1):

  • Comprehensive decision tracking system with full lifecycle management (record → analyze → query → precedent → influence)
  • Advanced KG algorithm integration: centrality analysis, community detection, node embeddings with ContextGraph
  • Enhanced AgentContext with granular feature flags for decision tracking, KG algorithms, and vector store features
  • PolicyException model replacing conflicting Exception name for meaningful business domain modeling
  • GraphStore validation preventing runtime failures with explicit capability checking
  • Hybrid search combining semantic, structural, and category similarity with configurable weights
  • Decision influence analysis with centrality measures and causal chain tracking
  • Policy management with versioning, compliance checking, and exception handling
  • Production-ready architecture with audit trails, security, and scalability features
  • 9 critical bug fixes: logging, security, audit trails, API compatibility, Cypher queries, centrality access, validation, naming
  • Comprehensive documentation with usage guides, production examples, and API references
  • 100% test coverage with all validation tests passing (9/9 tests)
  • Enterprise-grade features for financial services, healthcare, legal, and business domains
  • Complete backward compatibility with existing semantica components
  • Performance optimizations: caching, indexing, and efficient graph operations

• Added PgVector Store Support (PR #303 by @Sameer6305, @KaifAhmad1):

  • Native PostgreSQL vector storage using pgvector extension with full integration
  • Multiple distance metrics: cosine, L2/Euclidean, inner product with automatic score normalization
  • Advanced indexing: HNSW and IVFFlat for approximate nearest neighbor search with tunable parameters
  • JSONB metadata storage with flexible filtering capabilities and batch operations
  • Connection pooling support with psycopg3/psycopg2 fallback and efficient resource management
  • Comprehensive VectorStore integration with backend delegation and unified API
  • Idempotent index creation and table management with safe migration support
  • Production-ready security: SQL injection protection with psycopg_sql.SQL() and input validation
  • Performance optimizations: UUID4-based IDs, batch executemany operations, connection pooling
  • Full backward compatibility with existing vector store implementations
  • 36+ comprehensive test cases with Docker integration and dependency skipping
  • Complete documentation with setup guides, examples, and performance tuning
  • CI/CD integration: resolved benchmark compatibility and fixed documentation links

• Improved Vector Store for Decision Tracking (PR #293 by @KaifAhmad1):

  • Comprehensive decision tracking capabilities with hybrid search combining semantic and structural embeddings
  • New DecisionEmbeddingPipeline for generating semantic and structural embeddings with KG algorithm integration
  • HybridSimilarityCalculator with configurable weights (semantic: 0.7, structural: 0.3)
  • DecisionContext high-level interface for decision management with explainable AI features
  • ContextRetriever with hybrid precedent search and multi-hop reasoning
  • User-friendly convenience API: quick_decision(), find_precedents(), explain(), similar_to(), batch_decisions(), filter_decisions()
  • Knowledge Graph algorithm integration: Node2Vec, PathFinder, CommunityDetector, CentralityCalculator, SimilarityCalculator, ConnectivityAnalyzer
  • Explainable AI with path tracing, confidence scoring, and comprehensive decision explanations
  • Performance optimizations: 0.028s per decision processing, 0.031s search performance, ~0.8KB per decision memory usage
  • 100% backward compatibility maintained with existing VectorStore functionality
  • 34+ comprehensive tests covering all functionality including end-to-end scenarios and performance benchmarks
  • Real-world validation examples for banking and insurance domains
  • Documentation with clear imports, examples, and API references

• Improved Graph Algorithms in KG Module (PR #292 by @KaifAhmad1):

  • Complete algorithm suite with 30+ graph algorithms across 7 categories
  • Node Embeddings: Node2Vec, DeepWalk, Word2Vec for structural similarity analysis
  • Similarity Analysis: Cosine, Euclidean, Manhattan, Correlation metrics with batch processing
  • Path Finding: Dijkstra, A*, BFS, K-shortest paths for route and network analysis
  • Link Prediction: Preferential attachment, Jaccard, Adamic-Adar for network completion
  • Centrality Analysis: Degree, Betweenness, Closeness, PageRank for importance ranking
  • Community Detection: Louvain, Leiden, Label propagation for clustering analysis
  • Connectivity Analysis: Components, bridges, density for network robustness
  • Unified provenance tracking system with GraphBuilderWithProvenance and AlgorithmTrackerWithProvenance
  • Complete execution tracking with metadata, timestamps, and reproducibility IDs
  • Comprehensive test coverage with 5 test suites and 40+ test methods
  • Professional documentation overhaul for all modules and reference documentation
  • Enterprise-ready functionality with error handling and NetworkX compatibility
  • Performance optimizations with sparse matrix operations and batch processing
  • Full backward compatibility maintained with gradual migration support

• Improved Security Configuration with Dependabot:

  • Configured bi-weekly security updates with manual review by @KaifAhmad1
  • Implemented automated security scans (Monday & Thursday at 7 AM IST) with Bandit, Safety, Semgrep
  • Added security-critical package grouping (cryptography, requests, urllib3, certifi, pyopenssl)
  • Enterprise-grade security with audit trail, compliance features, and zero auto-merge
  • Optimized IST timezone scheduling (Security scans: 7 AM IST, PRs: 9 AM IST)
  • Aligned with new Dependabot features: open-source proxy support, smart dependency grouping for Snowflake/Arrow/benchmark features, private registry support, semantic commit prefixes, and latest GitHub security best practices

• ResourceScheduler Deadlock Fix and Performance Improvements (PR #299, #301 by @d4ndr4d3, @KaifAhmad1):

  • Fixed critical deadlock in ResourceScheduler by replacing threading.Lock() with threading.RLock()
  • Resolved nested lock acquisition issue in allocate_resources() → allocate_cpu/memory/gpu() calls
  • Added allocation validation with ValidationError when no resources can be allocated
  • Improved performance by moving progress tracking updates outside lock scope
  • Implemented comprehensive resource cleanup on allocation failures to prevent leaks
  • Added complete regression test suite (6 tests) for deadlock prevention and edge cases
  • Improved error handling and documentation for better operator visibility
  • Zero breaking changes, maintains thread safety and backward compatibility

[0.2.7] - 2026-02-09

Added / Changed

• Snowflake Connector for Data Ingestion (PR #276 by @Sameer6305):

  • Native Snowflake connector with multi-authentication (password, OAuth, key-pair, SSO)
  • Table and query ingestion with pagination, schema introspection, batch processing
  • SQL injection prevention via identifier escaping, OAuth token validation
  • Progress tracking integration, context manager support, document export
  • 24 comprehensive unit tests with mocking, complete documentation and examples
  • Added as optional dependency db-snowflake with snowflake-connector-python>=3.0.0

• Apache Arrow Export Support (PR #273 by @Sameer6305):

  • Added Apache Arrow exporter with explicit schemas, entity/relationship export, compression support
  • Integrated with export module and method registry, Pandas/DuckDB compatible
  • 20 unit tests + 1 integration test, complete documentation with examples

• Comprehensive Benchmark Suite with Regression CLI (PR #289 by @ZohaibHassan16, @KaifAhmad1):

  • 137+ benchmarks across all 10 Semantica modules (Input, Core, Storage, Context, QA, Ontology, etc.)
  • Environment-agnostic design with robust mocking system for CI/CD compatibility
  • Statistical regression detection using Z-score analysis with configurable thresholds
  • Automated performance auditing via GitHub Actions workflow
  • Comprehensive documentation suite (benchmarks.md, architecture guides, usage examples)
  • Zero breaking changes, production-ready with ultra-fast text processing (>10,000 ops/s)
  • Added benchmark runner CLI: python benchmarks/benchmark_runner.py

[0.2.6] - 2026-02-03

Added / Changed

• W3C PROV-O Compliant Provenance Tracking (#254, #246):

  • Comprehensive provenance tracking system with W3C PROV-O compliance across all 17 Semantica modules
  • Core Module: ProvenanceManager, W3C PROV-O schemas, storage backends (InMemory, SQLite), SHA-256 integrity verification
  • Module Integrations: Semantic Extract, LLMs (Groq, OpenAI, HuggingFace, LiteLLM), Pipeline, Context, Ingest, Embeddings, Graph/Vector/Triplet stores, Reasoning, Conflicts, Deduplication, Export, Parse, Normalize, Ontology, Visualization
  • Features: Complete lineage tracking (Document → Chunk → Entity → Relationship → Graph), LLM tracking (tokens, costs, latency), source tracking, bridge axioms for domain transformations
  • Compliance Infrastructure: W3C PROV-O, FDA 21 CFR Part 11, SOX, HIPAA, TNFD
  • Testing: 237 tests covering core functionality, all 17 module integrations, edge cases, backward compatibility
  • Design: Opt-in with provenance=False by default, zero breaking changes, no new dependencies
  • Contributed by @KaifAhmad1

• Enhanced Change Management Module (#248, #243):

  • Enterprise-grade version control for knowledge graphs and ontologies with persistent storage and audit trails
  • Core Classes: TemporalVersionManager (KG versioning), OntologyVersionManager (ontology versioning), ChangeLogEntry (metadata)
  • Storage: SQLite (persistent) and in-memory backends with thread-safe operations
  • Features: SHA-256 checksums, detailed entity/relationship diffs, structural ontology comparison, email validation
  • Compliance Infrastructure: HIPAA, SOX, FDA 21 CFR Part 11 with immutable audit trails
  • Testing: 104 tests (100% pass) - unit, integration, compliance, performance, edge cases
  • Performance: 17.6ms for 10k entities, 510+ ops/sec concurrent, handles 5k+ entity graphs
  • Migration: Backward compatible, simplified class names, zero external dependencies
  • Contributed by @KaifAhmad1

• CSV Ingestion Enhancements (PR #244 by @saloni0318)

  • Auto-detect CSV encoding (chardet) and delimiter (csv.Sniffer)
  • Tolerant decoding and malformed-row handling (on_bad_lines='warn')
  • Optional chunked reading for large files; metadata tracks detected values
  • Expanded unit tests covering delimiters, quoted/multiline fields, header overrides, chunks, and NaN preservation

• Tests: Comprehensive units for TextNormalizer (PR #242 by @ZohaibHassan16)

  • Added focused test coverage for TextNormalizer behavior across inputs

• Tests: Register integration mark and tidy ingest test warnings (PR #241 by @KaifAhmad1)

  • Introduced integration test marker and reduced noisy warnings in ingest tests

• Ingest Unit Tests (#239, #232):

  • Comprehensive unit tests for ingestion modules (file, web, and feed ingestors)
  • Coverage: File scanning (local/cloud S3/GCS/Azure), web ingestion (URL/sitemap/robots.txt), RSS/Atom feed parsing
  • Testing: 998 lines of test code with mocked external dependencies for fast, isolated execution
  • Results: file_ingestor (86%), web_ingestor (86%), feed_ingestor (80%) coverage
  • Covers happy paths, edge cases, and error handling
  • Contributed by @Mohammed2372

Fixed

• Temperature Compatibility Fix (#256, #252):

  • Fixed hardcoded temperature=0.3 that broke compatibility with models requiring specific temperature values (e.g., gpt-5-mini)
  • Added _add_if_set helper method to BaseProvider that only passes parameters when explicitly set
  • When temperature=None, parameter is omitted allowing APIs to use model defaults
  • Updated all 5 providers: OpenAI, Groq, Gemini, Ollama, DeepSeek
  • Reduced code by ~85 lines with cleaner parameter handling
  • Comprehensive test coverage added (10 temperature tests, all passing)
  • Backward compatible - no breaking changes
  • Contributed by @F0rt1s and @IGES-Institut

• JenaStore Empty Graph Bug (#257, #258):

  • Fixed ProcessingError: Graph not initialized when operating on empty (but initialized) graphs
  • Replaced implicit if not self.graph: checks with explicit if self.graph is None: validation in 5 methods (add_triplets, get_triplets, delete_triplet, execute_sparql, serialize)
  • Properly distinguishes None (uninitialized) from empty graphs (initialized with 0 triplets)
  • Unblocks benchmarking suite, fresh deployments, and testing workflows
  • Contributed by @ZohaibHassan16

[0.2.5] - 2026-01-27

Added

• Pinecone Vector Store Support:
  • Implemented native Pinecone support (PineconeStore) with full CRUD capabilities.
  • Added support for serverless and pod-based indexes, namespaces, and metadata filtering.
  • Integrated with VectorStore unified interface and registry.
  • (Closes #219, Resolves #220)
• Configurable LLM Retry Logic:
  • Exposed max_retries parameter in NERExtractor, RelationExtractor, TripletExtractor and low-level extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Defaults to 3 retries to prevent infinite loops during JSON validation failures or API timeouts.
  • Propagated retry configuration through chunked processing helpers to ensure consistent behavior for long documents.
  • Updated 03_Earnings_Call_Analysis.ipynb to use max_retries=3 by default.

Added

• Bring Your Own Model (BYOM) Support:
  • Enabled full support for custom Hugging Face models in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added support for custom tokenizers in HuggingFaceModelLoader to handle models with non-standard tokenization requirements.
  • Implemented robust fallback logic for model selection: runtime options (extract(model=...)) now correctly override configuration defaults.
• Enhanced NER Implementation:
  • Added configurable aggregation strategies (simple, first, average, max) to extract_entities_huggingface for better sub-word token handling.
  • Implemented robust IOB/BILOU parsing to reconstruct entities from raw model outputs when structured output is unavailable.
  • Added confidence scoring for aggregated entities.
• Relation Extraction Improvements:
  • Implemented standard entity marker technique (wrapping subject/object with <subj>, <obj> tags) in extract_relations_huggingface for compatibility with sequence classification models.
  • Added structured output parsing to convert raw model predictions into validated Relation objects.
• Triplet Extraction Completion:
  • Added specialized parsing for Seq2Seq models (e.g., REBEL) in extract_triplets_huggingface to generate structured triplets directly from text.
  • Implemented post-processing logic to clean and validate generated triplets.

Fixed

• LLM Extraction Stability:
  • Fixed infinite retry loops in BaseProvider by strictly enforcing max_retries limit during structured output generation.
  • Resolved stuck execution in earnings call analysis notebooks when using smaller models (e.g., Llama 3 8B) that frequently produce invalid JSON.
• Model Parameter Precedence:
  • Fixed issue where configuration defaults took precedence over runtime arguments in Hugging Face extractors. Runtime options now correctly override config values.
• Import Handling:
  • Fixed circular import issues in test suites by implementing robust mocking strategies.

[0.2.4] - 2026-01-22

Added

• Ontology Ingestion Module:
  • Implemented OntologyIngestor in semantica.ingest for parsing RDF/OWL files (Turtle, RDF/XML, JSON-LD, N3) into standardized OntologyData objects.
  • Added ingest_ontology convenience function and integrated it into the unified ingest(source_type="ontology") interface.
  • Added recursive directory scanning support for batch ontology ingestion.
  • Exposed ingestion tools in semantica.ontology for better discoverability.
  • Added OntologyData dataclass for consistent metadata handling (source path, format, timestamps).
• Documentation:
  • Ontology Usage Guide: Updated ontology_usage.md with comprehensive examples for single-file and directory ingestion.
  • API Reference: Updated ontology.md with OntologyIngestor class documentation and method details.
• Tests:
  • Comprehensive Test Suite: Added tests/ingest/test_ontology_ingestor.py covering all supported formats, error handling, and unified interface integration.
  • Demo Script: Added examples/demo_ontology_ingest.py for end-to-end usage demonstration.

[0.2.3] - 2026-01-20

Fixed

• LLM Relation Extraction Parsing:
  • Fixed relation extraction returning zero relations despite successful API calls to Groq and other providers
  • Normalized typed responses from instructor/OpenAI/Groq to consistent dict format before parsing
  • Added structured JSON fallback when typed generation yields zero relations to avoid silent empty outputs
  • Removed acceptance of extra kwargs (max_tokens, max_entities_prompt) from relation extraction internals
  • Filtered kwargs passed to provider LLM calls to only temperature and verbose
• API Parameter Handling:
  • Limited kwargs forwarded in chunked extraction helper to prevent parameter leakage
  • Ensured minimal, safe parameters are passed to provider calls
• Pipeline Circular Import (Issues #192, #193):
  • Fixed circular import between pipeline_builder and pipeline_validator triggered during semantica.pipeline import
  • Lazy-loaded PipelineValidator inside PipelineBuilder.__init__ and guarded type hints with TYPE_CHECKING
  • Ensured from semantica.deduplication import DuplicateDetector no longer fails even when pipeline module is imported
• JupyterLab Progress Output (Issue #181):
  • Added SEMANTICA_DISABLE_JUPYTER_PROGRESS environment variable to disable rich Jupyter/Colab progress tables
  • When enabled, progress falls back to console-style output, preventing infinite scrolling and JupyterLab out-of-memory errors

Added

• Comprehensive Test Suite:
• • Added unit tests (tests/test_relations_llm.py) with mocked LLM provider covering both typed and structured response paths
• • Added integration tests (tests/integration/test_relations_groq.py) for real Groq API calls with environment variable API key
• • Tests validate relation extraction completion and result parsing across different response formats
• Amazon Neptune Dev Environment:
• • Added CloudFormation template (cookbook/introduction/neptune-setup.yaml) to provision a dev Neptune cluster with public endpoint and IAM auth enabled
• • Documented deployment, cost estimates, and IAM User vs IAM Role best practices in cookbook/introduction/21_Amazon_Neptune_Store.ipynb
• • Added cfn-lint to .pre-commit-config.yaml for validating CloudFormation templates while excluding neptune-setup.yaml from generic YAML linters
• Vector Store High-Performance Ingestion:
• • Added VectorStore.add_documents for high-throughput ingestion with automatic embedding generation, batching, and parallel processing
• • Added VectorStore.embed_batch helper for generating embeddings for lists of texts without immediately storing them
• • Enabled default parallel ingestion in VectorStore with max_workers=6 for common workloads
• • Added dedicated documentation page docs/vector_store_usage.md describing high-performance vector store usage and configuration
• • Added tests/vector_store/test_vector_store_parallel.py covering parallel vs sequential performance, error handling, and edge cases for add_documents and embed_batch

Changed

• Relation Extraction API:
• • Simplified parameter interface by removing unused kwargs that were previously ignored
• • Improved error handling and verbose logging for debugging relation extraction issues
• • Enhanced robustness of post-response parsing across different LLM providers
• Vector Store Defaults and Examples:
• • Standardized VectorStore default concurrency to max_workers=6 for parallel ingestion
• • Updated vector store reference documentation and usage guides to rely on implicit defaults instead of requiring manual max_workers configuration in examples

[0.2.2] - 2026-01-15

Added

• Parallel Extraction Engine:
  • Implemented high-throughput parallel batch processing across all core extractors (NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticNetworkExtractor) using concurrent.futures.ThreadPoolExecutor.
  • Added max_workers configuration parameter (default: 1) to all extractor extract() methods, allowing users to tune concurrency based on available CPU cores or API rate limits.
  • Parallel Chunking: Implemented parallel processing for large document chunking in _extract_entities_chunked and _extract_relations_chunked, significantly reducing latency for long-form text analysis.
  • Thread-Safe Progress Tracking: Enhanced ProgressTracker to handle concurrent updates from multiple threads without race conditions during batch processing.
• Semantic Extract Performance & Regression:
  • Added edge-case regression suite covering max worker defaults, LLM prompt entity filtering, and extractor reuse.
  • Added a runnable real-use-case benchmark script for batch latency across NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticAnalyzer, and SemanticNetworkExtractor.
  • Added Groq LLM smoke tests that exercise LLM-based entities/relations/triplets when GROQ_API_KEY is available via environment configuration.

Security

• Credential Sanitization:
  • Removed hardcoded API keys from 8 cookbook notebooks to prevent secret leakage.
  • Enforced environment variable usage for GROQ_API_KEY across all examples.
• Secure Caching:
  • Updated ExtractionCache to exclude sensitive parameters (e.g., api_key, token, password) from cache key generation, preventing secret leakage and enabling safe cache sharing.
  • Upgraded cache key hashing algorithm from MD5 to SHA-256 for enhanced collision resistance and security.

Changed

• Gemini SDK Migration:
  • Migrated GeminiProvider to use the new google-genai SDK (v0.1.0+) to address deprecation warnings.
  • Implemented graceful fallback to google.generativeai for backward compatibility.
• Dependency Resolution:
  • Pinned opentelemetry-api and opentelemetry-sdk to 1.37.0 to resolve pip conflicts.
  • Updated protobuf and grpcio constraints for better stability.
• Entity Filtering Scope:
  • Removed entity filtering from non-LLM extraction flows to avoid accuracy regressions.
  • Applied entity downselection only to LLM relation prompt construction, while matching returned entities against the full original entity list.
• Batch Concurrency Defaults:
  • Standardized max_workers defaulting across semantic_extract and tuned for low-latency: ML-backed methods default to single-worker, while pattern/regex/rules/LLM/huggingface methods use a higher parallelism default capped by CPU.
  • Raised the global optimization.max_workers default to 8 for better throughput on batch workloads.

Performance

• Bottleneck Optimization (GitHub Issue #186):
  • Resolved Bottleneck #1 (Sequential Processing): Replaced sequential for loops with parallel execution for both document-level batches and intra-document chunks.
  • Performance Gains: Achieved ~1.89x speedup in real-world extraction scenarios (tested with Groq llama-3.3-70b-versatile on standard datasets).
  • Initialization Optimization: Refactored test suite to use class-level setUpClass for LLM provider initialization, eliminating redundant API client creation overhead.
• Low-Latency Entity Matching:
  • Avoided heavyweight embedding stack imports on common matches by improving fast matching heuristics and short-circuiting before embedding similarity.
  • Optimized entity matching to prioritize exact/substring/word-boundary matches and only fall back to embedding similarity when needed, reducing CPU overhead in LLM relation/triplet mapping.

[0.2.1] - 2026-01-12

Fixed

• LLM Output Stability (Bug #176):
  • Fixed incomplete JSON output issues by correctly propagating max_tokens parameter in extract_relations_llm.
  • Implemented automatic error handling that halves chunk sizes and retries when LLM context or output limits are exceeded.
  • Fixed AttributeError in provider integration by ensuring consistent parameter passing via **kwargs.
• Constraint Relaxations:
  • Removed hardcoded max_length constraints from Entity, Relation, and Triplet classes to support long-form semantic extraction (e.g., long descriptions or names).
• Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
• Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
• Fixed dependency compatibility issues by pinning protobuf>=5.29.1,<7.0 and grpcio>=1.71.2.
• Added missing dependencies GitPython and chardet to pyproject.toml.
• Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Changed

• Chunking Defaults:
  • Increased default max_text_length for auto-chunking to 64,000 characters (from 32k/16k) for OpenAI, Anthropic, Gemini, Groq, and DeepSeek providers.
  • Unified chunking logic across extract_entities_llm, extract_relations_llm, and extract_triplets_llm.
• Groq Support:
  • Standardized Groq provider defaults to use llama-3.3-70b-versatile with a 64k context window.
  • Added native support for max_tokens and max_completion_tokens to prevent output truncation.

Added

• Testing:
  • Added tests/reproduce_issue_176.py to validate max_tokens propagation and chunking behavior across all extractors.

[0.2.0] - 2026-01-10

Added

• Amazon Neptune Support:
  • Added AmazonNeptuneStore providing Amazon Neptune graph database integration via Bolt protocol and OpenCypher.
  • Implemented NeptuneAuthTokenManager extending Neo4j AuthManager for AWS IAM SigV4 signing with automatic token refresh.
  • Added robust connection handling: retry logic with backoff for transient errors (signature expired, connection closed) and driver recreation.
  • Added graph-amazon-neptune optional dependency group (boto3, neo4j).
  • Comprehensive test suite covering all GraphStore interface methods.
• Docling Integration:
  • Added DoclingParser in semantica.parse for high-fidelity document parsing using the Docling library.
  • Supports multi-format parsing (PDF, DOCX, PPTX, XLSX, HTML, images) with superior table extraction and structure understanding.
  • Implemented as a standalone parser supporting local execution, OCR, and multiple export formats (Markdown, HTML, JSON).
• Robust Extraction Fallbacks:
  • Implemented comprehensive fallback chains ("ML/LLM" -> "Pattern" -> "Last Resort") across NERExtractor, RelationExtractor, and TripletExtractor to prevent empty result lists.
  • Added "Last Resort" pattern matching in NERExtractor to identify capitalized words as generic entities when all other methods fail.
  • Added "Last Resort" adjacency-based relation extraction in RelationExtractor to create weak connections between adjacent entities if no relations are found.
  • Added fallback logic in TripletExtractor to convert relations to triplets or use rule-based extraction if standard methods fail.
• Provenance & Tracking:
  • Added count tracking to batch processing logs in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added batch_index and document_id to the metadata of all extracted entities, relations, triplets, semantic roles, and clusters for better traceability.
• Semantic Extract Improvements:
  • Introduced auto-chunking for long text processing in LLM extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Added silent_fail parameter to LLM extraction methods for configurable error handling.
  • Implemented robust JSON parsing and automatic retry logic (3 attempts with exponential backoff) in BaseProvider for all LLM providers.
  • Enhanced GroqProvider with better diagnostics and connectivity testing.
  • Added comprehensive entity, relation, and triplet deduplication for chunked extraction.
  • Added semantica/semantic_extract/schemas.py with canonical Pydantic models for consistent structured output.
• Testing:
  • Added comprehensive robustness test suite tests/semantic_extract/test_robustness_fallback.py for validating extraction fallbacks and metadata propagation.
  • Added comprehensive unit test suite tests/embeddings/test_model_switching.py for verifying dynamic model transitions and dimension updates.
  • Added end-to-end integration test suite for Knowledge Graph pipeline validation (GraphBuilder -> EntityResolver -> GraphAnalyzer).
• Other:
  • Added missing dependencies GitPython and chardet to pyproject.toml.
  • Robustified ID extraction across CentralityCalculator, CommunityDetector, and ConnectivityAnalyzer to handle various entity formats.
  • Improved Entity class hashability and equality logic in utils/types.py.

Changed

• Deduplication & Conflict Logic:
  • Removed internal deduplication logic from NERExtractor, RelationExtractor, and TripletExtractor.
  • Removed consistency/conflict checking from ExtractionValidator to defer to dedicated semantica/conflicts module.
  • Removed _deduplicate_* methods from semantica/semantic_extract/methods.py.
• Batch Processing & Consistency:
  • Standardized batch processing across all extractors (NERExtractor, RelationExtractor, TripletExtractor, SemanticNetworkExtractor, EventDetector, SemanticAnalyzer, CoreferenceResolver) using a unified extract/analyze/resolve method pattern with progress tracking.
  • Added provenance metadata (batch_index, document_id) to SemanticNetwork nodes/edges, Event objects, SemanticRole results, CoreferenceChain mentions, and SemanticCluster (tracking source document_ids).
  • Updated SemanticClusterer.cluster and SemanticAnalyzer.cluster_semantically to accept list of dictionaries (with content and id keys) for better document tracking during clustering.
  • Removed legacy check_triplet_consistency from TripletExtractor.
  • Removed validate_consistency and _check_consistency from ExtractionValidator.
• Weighted Scoring:
  • Clarified weighted confidence scoring (50% Method Confidence + 50% Type Similarity) in comments.
  • Explicitly labeled "Type Similarity" as "user-provided" in code comments to remove ambiguity.
• Refactoring:
  • Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
  • Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Fixed

• Critical Fixes:
  • Resolved NameError in extraction_validator.py by adding missing Union import.
  • Resolved issues where extractors would return empty lists for valid input text when primary extraction methods failed.
  • Fixed metadata initialization issue in batch processing where batch_index and document_id were occasionally missing from extracted items.
  • Ensured LLMExtraction methods (enhance_entities, enhance_relations) return original input instead of failing or returning empty results when LLM providers are unavailable.
• Component Fixes:
  • Fixed model switching bug in TextEmbedder where internal state was not cleared, preventing dynamic updates between fastembed and sentence_transformers (#160).
  • Implemented model-intrinsic embedding dimension detection in TextEmbedder to ensure consistency between models and vector databases.
  • Updated set_model to properly refresh configuration and dimensions during model switches.
  • Fixed TypeError: unhashable type: 'Entity' in GraphAnalyzer when processing graphs with raw Entity objects or dictionaries in relationships (#159).
  • Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
  • Fixed dependency compatibility issues by pinning protobuf==4.25.3 and grpcio==1.67.1.
  • Fixed a bug in TripletExtractor where the validate_triplets method was shadowed by an internal attribute.
  • Fixed incorrect TextSplitter import path in the semantic_extract.methods module.

[0.1.1] - 2026-01-05

Added

• Exported DoclingParser and DoclingMetadata from semantica.parse for easier access.
• Added comprehensive DoclingParser usage examples to README and documentation.
• Added Windows-specific troubleshooting note for PyTorch DLL issues.

Fixed

• Fixed DoclingParser import/export issues across platforms (Windows, Linux, Google Colab).
• Improved error messaging when optional docling dependency is missing.
• Fixed versioning inconsistencies across the framework.

[0.1.0] - 2025-12-31

Added

• New command-line interface (semantica CLI) with support for knowledge base building and info commands.
• Integrated FastAPI-based REST API server for remote access to framework functionality.
• Dedicated background worker component for scalable task processing and pipeline execution.
• Framework-level versioning configuration for PyPI distribution.
• Automated release workflow with Trusted Publishing support.

Changed

• Updated versioning across the framework to 0.1.0.
• Refined entry point configurations in pyproject.toml.
• Improved lazy module loading for core framework components.

[0.0.5] - 2025-11-26

Changed

• Configured Trusted Publishing for secure automated PyPI deployments

[0.0.4] - 2025-11-26

Changed

• Fixed PyPI deployment issues from v0.0.3

[0.0.3] - 2025-11-25

Changed

• Simplified CI/CD workflows - removed failing tests and strict linting
• Combined release and PyPI publishing into single workflow
• Simplified security scanning to weekly pip-audit only
• Streamlined GitHub Actions configuration

Added

• Comprehensive issue templates (Bug, Feature, Documentation, Support, Grant/Partnership)
• Updated pull request template with clear guidelines
• Community support documentation (SUPPORT.md)
• Funding and sponsorship configuration (FUNDING.yml)
• GitHub configuration README for maintainers
• 10+ new domain-specific cookbook examples (Finance, Healthcare, Cybersecurity, etc.)

Removed

• Redundant scripts folder (8 shell/PowerShell scripts)
• Unnecessary automation workflows (label-issues, mark-answered)
• Excessive issue templates

[0.0.2] - 2025-11-25

Changed

• Updated README with streamlined content and better examples
• Added more notebooks to cookbook
• Improved documentation structure

[0.0.1] - 2024-01-XX

Added

• Core framework architecture
• Universal data ingestion (multiple file formats)
• Semantic intelligence engine (NER, relation extraction, event detection)
• Knowledge graph construction with entity resolution
• 6-stage ontology generation pipeline
• GraphRAG engine for hybrid retrieval
• Multi-agent system infrastructure
• Production-ready quality assurance modules
• Comprehensive documentation with MkDocs
• Cookbook with interactive tutorials
• Support for multiple vector stores (Weaviate, Qdrant, FAISS)
• Support for multiple graph databases (Neo4j, NetworkX, RDFLib)
• Temporal knowledge graph support
• Conflict detection and resolution
• Deduplication and entity merging
• Schema template enforcement
• Seed data management
• Multi-format export (RDF, JSON-LD, CSV, GraphML)
• Visualization tools
• Pipeline orchestration
• Streaming support (Kafka, RabbitMQ, Kinesis)
• Context engineering for AI agents
• Reasoning and inference engine

Documentation

• Getting started guide
• API reference for all modules
• Concepts and architecture documentation
• Use case examples
• Cookbook tutorials
• Community projects showcase

---

Types of Changes

• Added for new features
• Changed for changes in existing functionality
• Deprecated for soon-to-be removed features
• Removed for now removed features
• Fixed for any bug fixes
• Security for vulnerability fixes

Migration Guides

When breaking changes are introduced, migration guides will be provided in the release notes and documentation.

---

For detailed release notes, see GitHub Releases.

Legacy Changelog Snapshot A (Preserved Merge Artifact)

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

• OWLGenerator user-facing schema compatibility fixes (Issue #446):

  • Fixed OWL class/property IRI identifier fallback order to prefer label and then name.
  • Fixed datatype property handling to accept scalar and list range values in rdflib path (including xsd:*, full IRIs, and local names), preventing list-based .startswith() crashes.
  • Fixed generated class/property/domain/range IRIs to use the current ontology dict uri namespace for each generation call (instead of drifting to default namespace manager base URI when per-entity uri is omitted).
  • Fixed subClassOf / subclassOf parent resolution so local class names are expanded to ontology IRIs consistently with domain/range behavior.
  • Added/expanded regression coverage in ests/ontology/test_ontology_comprehensive.py ( est_owl_generator_user_facing_schema_compatibility) for label-first fallback, lowercase subclassOf, datatype range lists, and ontology namespace consistency.

• Fixed: PolicyEngine latest version selection on ContextGraph; AgentContext fallback robustness and secure logging (PR #TBD by @KaifAhmad1)

• Tests: Added ContextGraph fallback and AgentContext smoke tests; full suite passing

• Context Engineering Enhancement (PR #307 by @KaifAhmad1):

  • Comprehensive decision tracking system with full lifecycle management (record → analyze → query → precedent → influence)
  • Advanced KG algorithm integration: centrality analysis, community detection, node embeddings with ContextGraph
  • Enhanced AgentContext with granular feature flags for decision tracking, KG algorithms, and vector store features
  • PolicyException model replacing conflicting Exception name for meaningful business domain modeling
  • GraphStore validation preventing runtime failures with explicit capability checking
  • Hybrid search combining semantic, structural, and category similarity with configurable weights
  • Decision influence analysis with centrality measures and causal chain tracking
  • Policy management with versioning, compliance checking, and exception handling
  • Production-ready architecture with audit trails, security, and scalability features
  • 9 critical bug fixes: logging, security, audit trails, API compatibility, Cypher queries, centrality access, validation, naming
  • Comprehensive documentation with usage guides, production examples, and API references
  • 100% test coverage with all validation tests passing (9/9 tests)
  • Enterprise-grade features for financial services, healthcare, legal, and business domains
  • Complete backward compatibility with existing semantica components
  • Performance optimizations: caching, indexing, and efficient graph operations

• Added PgVector Store Support (PR #303 by @Sameer6305, @KaifAhmad1):

  • Native PostgreSQL vector storage using pgvector extension with full integration
  • Multiple distance metrics: cosine, L2/Euclidean, inner product with automatic score normalization
  • Advanced indexing: HNSW and IVFFlat for approximate nearest neighbor search with tunable parameters
  • JSONB metadata storage with flexible filtering capabilities and batch operations
  • Connection pooling support with psycopg3/psycopg2 fallback and efficient resource management
  • Comprehensive VectorStore integration with backend delegation and unified API
  • Idempotent index creation and table management with safe migration support
  • Production-ready security: SQL injection protection with psycopg_sql.SQL() and input validation
  • Performance optimizations: UUID4-based IDs, batch executemany operations, connection pooling
  • Full backward compatibility with existing vector store implementations
  • 36+ comprehensive test cases with Docker integration and dependency skipping
  • Complete documentation with setup guides, examples, and performance tuning
  • CI/CD integration: resolved benchmark compatibility and fixed documentation links

• Improved Vector Store for Decision Tracking (PR #293 by @KaifAhmad1):

  • Comprehensive decision tracking capabilities with hybrid search combining semantic and structural embeddings
  • New DecisionEmbeddingPipeline for generating semantic and structural embeddings with KG algorithm integration
  • HybridSimilarityCalculator with configurable weights (semantic: 0.7, structural: 0.3)
  • DecisionContext high-level interface for decision management with explainable AI features
  • ContextRetriever with hybrid precedent search and multi-hop reasoning
  • User-friendly convenience API: quick_decision(), find_precedents(), explain(), similar_to(), batch_decisions(), filter_decisions()
  • Knowledge Graph algorithm integration: Node2Vec, PathFinder, CommunityDetector, CentralityCalculator, SimilarityCalculator, ConnectivityAnalyzer
  • Explainable AI with path tracing, confidence scoring, and comprehensive decision explanations
  • Performance optimizations: 0.028s per decision processing, 0.031s search performance, ~0.8KB per decision memory usage
  • 100% backward compatibility maintained with existing VectorStore functionality
  • 34+ comprehensive tests covering all functionality including end-to-end scenarios and performance benchmarks
  • Real-world validation examples for banking and insurance domains
  • Documentation with clear imports, examples, and API references

• Improved Graph Algorithms in KG Module (PR #292 by @KaifAhmad1):

  • Complete algorithm suite with 30+ graph algorithms across 7 categories
  • Node Embeddings: Node2Vec, DeepWalk, Word2Vec for structural similarity analysis
  • Similarity Analysis: Cosine, Euclidean, Manhattan, Correlation metrics with batch processing
  • Path Finding: Dijkstra, A*, BFS, K-shortest paths for route and network analysis
  • Link Prediction: Preferential attachment, Jaccard, Adamic-Adar for network completion
  • Centrality Analysis: Degree, Betweenness, Closeness, PageRank for importance ranking
  • Community Detection: Louvain, Leiden, Label propagation for clustering analysis
  • Connectivity Analysis: Components, bridges, density for network robustness
  • Unified provenance tracking system with GraphBuilderWithProvenance and AlgorithmTrackerWithProvenance
  • Complete execution tracking with metadata, timestamps, and reproducibility IDs
  • Comprehensive test coverage with 5 test suites and 40+ test methods
  • Professional documentation overhaul for all modules and reference documentation
  • Enterprise-ready functionality with error handling and NetworkX compatibility
  • Performance optimizations with sparse matrix operations and batch processing
  • Full backward compatibility maintained with gradual migration support

• Improved Security Configuration with Dependabot:

  • Configured bi-weekly security updates with manual review by @KaifAhmad1
  • Implemented automated security scans (Monday & Thursday at 7 AM IST) with Bandit, Safety, Semgrep
  • Added security-critical package grouping (cryptography, requests, urllib3, certifi, pyopenssl)
  • Enterprise-grade security with audit trail, compliance features, and zero auto-merge
  • Optimized IST timezone scheduling (Security scans: 7 AM IST, PRs: 9 AM IST)
  • Aligned with new Dependabot features: open-source proxy support, smart dependency grouping for Snowflake/Arrow/benchmark features, private registry support, semantic commit prefixes, and latest GitHub security best practices

• ResourceScheduler Deadlock Fix and Performance Improvements (PR #299, #301 by @d4ndr4d3, @KaifAhmad1):

  • Fixed critical deadlock in ResourceScheduler by replacing threading.Lock() with threading.RLock()
  • Resolved nested lock acquisition issue in allocate_resources() → allocate_cpu/memory/gpu() calls
  • Added allocation validation with ValidationError when no resources can be allocated
  • Improved performance by moving progress tracking updates outside lock scope
  • Implemented comprehensive resource cleanup on allocation failures to prevent leaks
  • Added complete regression test suite (6 tests) for deadlock prevention and edge cases
  • Improved error handling and documentation for better operator visibility
  • Zero breaking changes, maintains thread safety and backward compatibility

[0.2.7] - 2026-02-09

Added / Changed

• Snowflake Connector for Data Ingestion (PR #276 by @Sameer6305):

  • Native Snowflake connector with multi-authentication (password, OAuth, key-pair, SSO)
  • Table and query ingestion with pagination, schema introspection, batch processing
  • SQL injection prevention via identifier escaping, OAuth token validation
  • Progress tracking integration, context manager support, document export
  • 24 comprehensive unit tests with mocking, complete documentation and examples
  • Added as optional dependency db-snowflake with snowflake-connector-python>=3.0.0

• Apache Arrow Export Support (PR #273 by @Sameer6305):

  • Added Apache Arrow exporter with explicit schemas, entity/relationship export, compression support
  • Integrated with export module and method registry, Pandas/DuckDB compatible
  • 20 unit tests + 1 integration test, complete documentation with examples

• Comprehensive Benchmark Suite with Regression CLI (PR #289 by @ZohaibHassan16, @KaifAhmad1):

  • 137+ benchmarks across all 10 Semantica modules (Input, Core, Storage, Context, QA, Ontology, etc.)
  • Environment-agnostic design with robust mocking system for CI/CD compatibility
  • Statistical regression detection using Z-score analysis with configurable thresholds
  • Automated performance auditing via GitHub Actions workflow
  • Comprehensive documentation suite (benchmarks.md, architecture guides, usage examples)
  • Zero breaking changes, production-ready with ultra-fast text processing (>10,000 ops/s)
  • Added benchmark runner CLI: python benchmarks/benchmark_runner.py

[0.2.6] - 2026-02-03

Added / Changed

• W3C PROV-O Compliant Provenance Tracking (#254, #246):

  • Comprehensive provenance tracking system with W3C PROV-O compliance across all 17 Semantica modules
  • Core Module: ProvenanceManager, W3C PROV-O schemas, storage backends (InMemory, SQLite), SHA-256 integrity verification
  • Module Integrations: Semantic Extract, LLMs (Groq, OpenAI, HuggingFace, LiteLLM), Pipeline, Context, Ingest, Embeddings, Graph/Vector/Triplet stores, Reasoning, Conflicts, Deduplication, Export, Parse, Normalize, Ontology, Visualization
  • Features: Complete lineage tracking (Document → Chunk → Entity → Relationship → Graph), LLM tracking (tokens, costs, latency), source tracking, bridge axioms for domain transformations
  • Compliance Infrastructure: W3C PROV-O, FDA 21 CFR Part 11, SOX, HIPAA, TNFD
  • Testing: 237 tests covering core functionality, all 17 module integrations, edge cases, backward compatibility
  • Design: Opt-in with provenance=False by default, zero breaking changes, no new dependencies
  • Contributed by @KaifAhmad1

• Enhanced Change Management Module (#248, #243):

  • Enterprise-grade version control for knowledge graphs and ontologies with persistent storage and audit trails
  • Core Classes: TemporalVersionManager (KG versioning), OntologyVersionManager (ontology versioning), ChangeLogEntry (metadata)
  • Storage: SQLite (persistent) and in-memory backends with thread-safe operations
  • Features: SHA-256 checksums, detailed entity/relationship diffs, structural ontology comparison, email validation
  • Compliance Infrastructure: HIPAA, SOX, FDA 21 CFR Part 11 with immutable audit trails
  • Testing: 104 tests (100% pass) - unit, integration, compliance, performance, edge cases
  • Performance: 17.6ms for 10k entities, 510+ ops/sec concurrent, handles 5k+ entity graphs
  • Migration: Backward compatible, simplified class names, zero external dependencies
  • Contributed by @KaifAhmad1

• CSV Ingestion Enhancements (PR #244 by @saloni0318)

  • Auto-detect CSV encoding (chardet) and delimiter (csv.Sniffer)
  • Tolerant decoding and malformed-row handling (on_bad_lines='warn')
  • Optional chunked reading for large files; metadata tracks detected values
  • Expanded unit tests covering delimiters, quoted/multiline fields, header overrides, chunks, and NaN preservation

• Tests: Comprehensive units for TextNormalizer (PR #242 by @ZohaibHassan16)

  • Added focused test coverage for TextNormalizer behavior across inputs

• Tests: Register integration mark and tidy ingest test warnings (PR #241 by @KaifAhmad1)

  • Introduced integration test marker and reduced noisy warnings in ingest tests

• Ingest Unit Tests (#239, #232):

  • Comprehensive unit tests for ingestion modules (file, web, and feed ingestors)
  • Coverage: File scanning (local/cloud S3/GCS/Azure), web ingestion (URL/sitemap/robots.txt), RSS/Atom feed parsing
  • Testing: 998 lines of test code with mocked external dependencies for fast, isolated execution
  • Results: file_ingestor (86%), web_ingestor (86%), feed_ingestor (80%) coverage
  • Covers happy paths, edge cases, and error handling
  • Contributed by @Mohammed2372

Fixed

• Temperature Compatibility Fix (#256, #252):

  • Fixed hardcoded temperature=0.3 that broke compatibility with models requiring specific temperature values (e.g., gpt-5-mini)
  • Added _add_if_set helper method to BaseProvider that only passes parameters when explicitly set
  • When temperature=None, parameter is omitted allowing APIs to use model defaults
  • Updated all 5 providers: OpenAI, Groq, Gemini, Ollama, DeepSeek
  • Reduced code by ~85 lines with cleaner parameter handling
  • Comprehensive test coverage added (10 temperature tests, all passing)
  • Backward compatible - no breaking changes
  • Contributed by @F0rt1s and @IGES-Institut

• JenaStore Empty Graph Bug (#257, #258):

  • Fixed ProcessingError: Graph not initialized when operating on empty (but initialized) graphs
  • Replaced implicit if not self.graph: checks with explicit if self.graph is None: validation in 5 methods (add_triplets, get_triplets, delete_triplet, execute_sparql, serialize)
  • Properly distinguishes None (uninitialized) from empty graphs (initialized with 0 triplets)
  • Unblocks benchmarking suite, fresh deployments, and testing workflows
  • Contributed by @ZohaibHassan16

[0.2.5] - 2026-01-27

Added

• Pinecone Vector Store Support:
  • Implemented native Pinecone support (PineconeStore) with full CRUD capabilities.
  • Added support for serverless and pod-based indexes, namespaces, and metadata filtering.
  • Integrated with VectorStore unified interface and registry.
  • (Closes #219, Resolves #220)
• Configurable LLM Retry Logic:
  • Exposed max_retries parameter in NERExtractor, RelationExtractor, TripletExtractor and low-level extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Defaults to 3 retries to prevent infinite loops during JSON validation failures or API timeouts.
  • Propagated retry configuration through chunked processing helpers to ensure consistent behavior for long documents.
  • Updated 03_Earnings_Call_Analysis.ipynb to use max_retries=3 by default.

Added

• Bring Your Own Model (BYOM) Support:
  • Enabled full support for custom Hugging Face models in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added support for custom tokenizers in HuggingFaceModelLoader to handle models with non-standard tokenization requirements.
  • Implemented robust fallback logic for model selection: runtime options (extract(model=...)) now correctly override configuration defaults.
• Enhanced NER Implementation:
  • Added configurable aggregation strategies (simple, first, average, max) to extract_entities_huggingface for better sub-word token handling.
  • Implemented robust IOB/BILOU parsing to reconstruct entities from raw model outputs when structured output is unavailable.
  • Added confidence scoring for aggregated entities.
• Relation Extraction Improvements:
  • Implemented standard entity marker technique (wrapping subject/object with <subj>, <obj> tags) in extract_relations_huggingface for compatibility with sequence classification models.
  • Added structured output parsing to convert raw model predictions into validated Relation objects.
• Triplet Extraction Completion:
  • Added specialized parsing for Seq2Seq models (e.g., REBEL) in extract_triplets_huggingface to generate structured triplets directly from text.
  • Implemented post-processing logic to clean and validate generated triplets.

Fixed

• LLM Extraction Stability:
  • Fixed infinite retry loops in BaseProvider by strictly enforcing max_retries limit during structured output generation.
  • Resolved stuck execution in earnings call analysis notebooks when using smaller models (e.g., Llama 3 8B) that frequently produce invalid JSON.
• Model Parameter Precedence:
  • Fixed issue where configuration defaults took precedence over runtime arguments in Hugging Face extractors. Runtime options now correctly override config values.
• Import Handling:
  • Fixed circular import issues in test suites by implementing robust mocking strategies.

[0.2.4] - 2026-01-22

Added

• Ontology Ingestion Module:
  • Implemented OntologyIngestor in semantica.ingest for parsing RDF/OWL files (Turtle, RDF/XML, JSON-LD, N3) into standardized OntologyData objects.
  • Added ingest_ontology convenience function and integrated it into the unified ingest(source_type="ontology") interface.
  • Added recursive directory scanning support for batch ontology ingestion.
  • Exposed ingestion tools in semantica.ontology for better discoverability.
  • Added OntologyData dataclass for consistent metadata handling (source path, format, timestamps).
• Documentation:
  • Ontology Usage Guide: Updated ontology_usage.md with comprehensive examples for single-file and directory ingestion.
  • API Reference: Updated ontology.md with OntologyIngestor class documentation and method details.
• Tests:
  • Comprehensive Test Suite: Added tests/ingest/test_ontology_ingestor.py covering all supported formats, error handling, and unified interface integration.
  • Demo Script: Added examples/demo_ontology_ingest.py for end-to-end usage demonstration.

[0.2.3] - 2026-01-20

Fixed

• LLM Relation Extraction Parsing:
  • Fixed relation extraction returning zero relations despite successful API calls to Groq and other providers
  • Normalized typed responses from instructor/OpenAI/Groq to consistent dict format before parsing
  • Added structured JSON fallback when typed generation yields zero relations to avoid silent empty outputs
  • Removed acceptance of extra kwargs (max_tokens, max_entities_prompt) from relation extraction internals
  • Filtered kwargs passed to provider LLM calls to only temperature and verbose
• API Parameter Handling:
  • Limited kwargs forwarded in chunked extraction helper to prevent parameter leakage
  • Ensured minimal, safe parameters are passed to provider calls
• Pipeline Circular Import (Issues #192, #193):
  • Fixed circular import between pipeline_builder and pipeline_validator triggered during semantica.pipeline import
  • Lazy-loaded PipelineValidator inside PipelineBuilder.__init__ and guarded type hints with TYPE_CHECKING
  • Ensured from semantica.deduplication import DuplicateDetector no longer fails even when pipeline module is imported
• JupyterLab Progress Output (Issue #181):
  • Added SEMANTICA_DISABLE_JUPYTER_PROGRESS environment variable to disable rich Jupyter/Colab progress tables
  • When enabled, progress falls back to console-style output, preventing infinite scrolling and JupyterLab out-of-memory errors

Added

• Comprehensive Test Suite:
• • Added unit tests (tests/test_relations_llm.py) with mocked LLM provider covering both typed and structured response paths
• • Added integration tests (tests/integration/test_relations_groq.py) for real Groq API calls with environment variable API key
• • Tests validate relation extraction completion and result parsing across different response formats
• Amazon Neptune Dev Environment:
• • Added CloudFormation template (cookbook/introduction/neptune-setup.yaml) to provision a dev Neptune cluster with public endpoint and IAM auth enabled
• • Documented deployment, cost estimates, and IAM User vs IAM Role best practices in cookbook/introduction/21_Amazon_Neptune_Store.ipynb
• • Added cfn-lint to .pre-commit-config.yaml for validating CloudFormation templates while excluding neptune-setup.yaml from generic YAML linters
• Vector Store High-Performance Ingestion:
• • Added VectorStore.add_documents for high-throughput ingestion with automatic embedding generation, batching, and parallel processing
• • Added VectorStore.embed_batch helper for generating embeddings for lists of texts without immediately storing them
• • Enabled default parallel ingestion in VectorStore with max_workers=6 for common workloads
• • Added dedicated documentation page docs/vector_store_usage.md describing high-performance vector store usage and configuration
• • Added tests/vector_store/test_vector_store_parallel.py covering parallel vs sequential performance, error handling, and edge cases for add_documents and embed_batch

Changed

• Relation Extraction API:
• • Simplified parameter interface by removing unused kwargs that were previously ignored
• • Improved error handling and verbose logging for debugging relation extraction issues
• • Enhanced robustness of post-response parsing across different LLM providers
• Vector Store Defaults and Examples:
• • Standardized VectorStore default concurrency to max_workers=6 for parallel ingestion
• • Updated vector store reference documentation and usage guides to rely on implicit defaults instead of requiring manual max_workers configuration in examples

[0.2.2] - 2026-01-15

Added

• Parallel Extraction Engine:
  • Implemented high-throughput parallel batch processing across all core extractors (NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticNetworkExtractor) using concurrent.futures.ThreadPoolExecutor.
  • Added max_workers configuration parameter (default: 1) to all extractor extract() methods, allowing users to tune concurrency based on available CPU cores or API rate limits.
  • Parallel Chunking: Implemented parallel processing for large document chunking in _extract_entities_chunked and _extract_relations_chunked, significantly reducing latency for long-form text analysis.
  • Thread-Safe Progress Tracking: Enhanced ProgressTracker to handle concurrent updates from multiple threads without race conditions during batch processing.
• Semantic Extract Performance & Regression:
  • Added edge-case regression suite covering max worker defaults, LLM prompt entity filtering, and extractor reuse.
  • Added a runnable real-use-case benchmark script for batch latency across NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticAnalyzer, and SemanticNetworkExtractor.
  • Added Groq LLM smoke tests that exercise LLM-based entities/relations/triplets when GROQ_API_KEY is available via environment configuration.

Security

• Credential Sanitization:
  • Removed hardcoded API keys from 8 cookbook notebooks to prevent secret leakage.
  • Enforced environment variable usage for GROQ_API_KEY across all examples.
• Secure Caching:
  • Updated ExtractionCache to exclude sensitive parameters (e.g., api_key, token, password) from cache key generation, preventing secret leakage and enabling safe cache sharing.
  • Upgraded cache key hashing algorithm from MD5 to SHA-256 for enhanced collision resistance and security.

Changed

• Gemini SDK Migration:
  • Migrated GeminiProvider to use the new google-genai SDK (v0.1.0+) to address deprecation warnings.
  • Implemented graceful fallback to google.generativeai for backward compatibility.
• Dependency Resolution:
  • Pinned opentelemetry-api and opentelemetry-sdk to 1.37.0 to resolve pip conflicts.
  • Updated protobuf and grpcio constraints for better stability.
• Entity Filtering Scope:
  • Removed entity filtering from non-LLM extraction flows to avoid accuracy regressions.
  • Applied entity downselection only to LLM relation prompt construction, while matching returned entities against the full original entity list.
• Batch Concurrency Defaults:
  • Standardized max_workers defaulting across semantic_extract and tuned for low-latency: ML-backed methods default to single-worker, while pattern/regex/rules/LLM/huggingface methods use a higher parallelism default capped by CPU.
  • Raised the global optimization.max_workers default to 8 for better throughput on batch workloads.

Performance

• Bottleneck Optimization (GitHub Issue #186):
  • Resolved Bottleneck #1 (Sequential Processing): Replaced sequential for loops with parallel execution for both document-level batches and intra-document chunks.
  • Performance Gains: Achieved ~1.89x speedup in real-world extraction scenarios (tested with Groq llama-3.3-70b-versatile on standard datasets).
  • Initialization Optimization: Refactored test suite to use class-level setUpClass for LLM provider initialization, eliminating redundant API client creation overhead.
• Low-Latency Entity Matching:
  • Avoided heavyweight embedding stack imports on common matches by improving fast matching heuristics and short-circuiting before embedding similarity.
  • Optimized entity matching to prioritize exact/substring/word-boundary matches and only fall back to embedding similarity when needed, reducing CPU overhead in LLM relation/triplet mapping.

[0.2.1] - 2026-01-12

Fixed

• LLM Output Stability (Bug #176):
  • Fixed incomplete JSON output issues by correctly propagating max_tokens parameter in extract_relations_llm.
  • Implemented automatic error handling that halves chunk sizes and retries when LLM context or output limits are exceeded.
  • Fixed AttributeError in provider integration by ensuring consistent parameter passing via **kwargs.
• Constraint Relaxations:
  • Removed hardcoded max_length constraints from Entity, Relation, and Triplet classes to support long-form semantic extraction (e.g., long descriptions or names).
• Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
• Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
• Fixed dependency compatibility issues by pinning protobuf>=5.29.1,<7.0 and grpcio>=1.71.2.
• Added missing dependencies GitPython and chardet to pyproject.toml.
• Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Changed

• Chunking Defaults:
  • Increased default max_text_length for auto-chunking to 64,000 characters (from 32k/16k) for OpenAI, Anthropic, Gemini, Groq, and DeepSeek providers.
  • Unified chunking logic across extract_entities_llm, extract_relations_llm, and extract_triplets_llm.
• Groq Support:
  • Standardized Groq provider defaults to use llama-3.3-70b-versatile with a 64k context window.
  • Added native support for max_tokens and max_completion_tokens to prevent output truncation.

Added

• Testing:
  • Added tests/reproduce_issue_176.py to validate max_tokens propagation and chunking behavior across all extractors.

[0.2.0] - 2026-01-10

Added

• Amazon Neptune Support:
  • Added AmazonNeptuneStore providing Amazon Neptune graph database integration via Bolt protocol and OpenCypher.
  • Implemented NeptuneAuthTokenManager extending Neo4j AuthManager for AWS IAM SigV4 signing with automatic token refresh.
  • Added robust connection handling: retry logic with backoff for transient errors (signature expired, connection closed) and driver recreation.
  • Added graph-amazon-neptune optional dependency group (boto3, neo4j).
  • Comprehensive test suite covering all GraphStore interface methods.
• Docling Integration:
  • Added DoclingParser in semantica.parse for high-fidelity document parsing using the Docling library.
  • Supports multi-format parsing (PDF, DOCX, PPTX, XLSX, HTML, images) with superior table extraction and structure understanding.
  • Implemented as a standalone parser supporting local execution, OCR, and multiple export formats (Markdown, HTML, JSON).
• Robust Extraction Fallbacks:
  • Implemented comprehensive fallback chains ("ML/LLM" -> "Pattern" -> "Last Resort") across NERExtractor, RelationExtractor, and TripletExtractor to prevent empty result lists.
  • Added "Last Resort" pattern matching in NERExtractor to identify capitalized words as generic entities when all other methods fail.
  • Added "Last Resort" adjacency-based relation extraction in RelationExtractor to create weak connections between adjacent entities if no relations are found.
  • Added fallback logic in TripletExtractor to convert relations to triplets or use rule-based extraction if standard methods fail.
• Provenance & Tracking:
  • Added count tracking to batch processing logs in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added batch_index and document_id to the metadata of all extracted entities, relations, triplets, semantic roles, and clusters for better traceability.
• Semantic Extract Improvements:
  • Introduced auto-chunking for long text processing in LLM extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Added silent_fail parameter to LLM extraction methods for configurable error handling.
  • Implemented robust JSON parsing and automatic retry logic (3 attempts with exponential backoff) in BaseProvider for all LLM providers.
  • Enhanced GroqProvider with better diagnostics and connectivity testing.
  • Added comprehensive entity, relation, and triplet deduplication for chunked extraction.
  • Added semantica/semantic_extract/schemas.py with canonical Pydantic models for consistent structured output.
• Testing:
  • Added comprehensive robustness test suite tests/semantic_extract/test_robustness_fallback.py for validating extraction fallbacks and metadata propagation.
  • Added comprehensive unit test suite tests/embeddings/test_model_switching.py for verifying dynamic model transitions and dimension updates.
  • Added end-to-end integration test suite for Knowledge Graph pipeline validation (GraphBuilder -> EntityResolver -> GraphAnalyzer).
• Other:
  • Added missing dependencies GitPython and chardet to pyproject.toml.
  • Robustified ID extraction across CentralityCalculator, CommunityDetector, and ConnectivityAnalyzer to handle various entity formats.
  • Improved Entity class hashability and equality logic in utils/types.py.

Changed

• Deduplication & Conflict Logic:
  • Removed internal deduplication logic from NERExtractor, RelationExtractor, and TripletExtractor.
  • Removed consistency/conflict checking from ExtractionValidator to defer to dedicated semantica/conflicts module.
  • Removed _deduplicate_* methods from semantica/semantic_extract/methods.py.
• Batch Processing & Consistency:
  • Standardized batch processing across all extractors (NERExtractor, RelationExtractor, TripletExtractor, SemanticNetworkExtractor, EventDetector, SemanticAnalyzer, CoreferenceResolver) using a unified extract/analyze/resolve method pattern with progress tracking.
  • Added provenance metadata (batch_index, document_id) to SemanticNetwork nodes/edges, Event objects, SemanticRole results, CoreferenceChain mentions, and SemanticCluster (tracking source document_ids).
  • Updated SemanticClusterer.cluster and SemanticAnalyzer.cluster_semantically to accept list of dictionaries (with content and id keys) for better document tracking during clustering.
  • Removed legacy check_triplet_consistency from TripletExtractor.
  • Removed validate_consistency and _check_consistency from ExtractionValidator.
• Weighted Scoring:
  • Clarified weighted confidence scoring (50% Method Confidence + 50% Type Similarity) in comments.
  • Explicitly labeled "Type Similarity" as "user-provided" in code comments to remove ambiguity.
• Refactoring:
  • Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
  • Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Fixed

• Critical Fixes:
  • Resolved NameError in extraction_validator.py by adding missing Union import.
  • Resolved issues where extractors would return empty lists for valid input text when primary extraction methods failed.
  • Fixed metadata initialization issue in batch processing where batch_index and document_id were occasionally missing from extracted items.
  • Ensured LLMExtraction methods (enhance_entities, enhance_relations) return original input instead of failing or returning empty results when LLM providers are unavailable.
• Component Fixes:
  • Fixed model switching bug in TextEmbedder where internal state was not cleared, preventing dynamic updates between fastembed and sentence_transformers (#160).
  • Implemented model-intrinsic embedding dimension detection in TextEmbedder to ensure consistency between models and vector databases.
  • Updated set_model to properly refresh configuration and dimensions during model switches.
  • Fixed TypeError: unhashable type: 'Entity' in GraphAnalyzer when processing graphs with raw Entity objects or dictionaries in relationships (#159).
  • Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
  • Fixed dependency compatibility issues by pinning protobuf==4.25.3 and grpcio==1.67.1.
  • Fixed a bug in TripletExtractor where the validate_triplets method was shadowed by an internal attribute.
  • Fixed incorrect TextSplitter import path in the semantic_extract.methods module.

[0.1.1] - 2026-01-05

Added

• Exported DoclingParser and DoclingMetadata from semantica.parse for easier access.
• Added comprehensive DoclingParser usage examples to README and documentation.
• Added Windows-specific troubleshooting note for PyTorch DLL issues.

Fixed

• Fixed DoclingParser import/export issues across platforms (Windows, Linux, Google Colab).
• Improved error messaging when optional docling dependency is missing.
• Fixed versioning inconsistencies across the framework.

[0.1.0] - 2025-12-31

Added

• New command-line interface (semantica CLI) with support for knowledge base building and info commands.
• Integrated FastAPI-based REST API server for remote access to framework functionality.
• Dedicated background worker component for scalable task processing and pipeline execution.
• Framework-level versioning configuration for PyPI distribution.
• Automated release workflow with Trusted Publishing support.

Changed

• Updated versioning across the framework to 0.1.0.
• Refined entry point configurations in pyproject.toml.
• Improved lazy module loading for core framework components.

[0.0.5] - 2025-11-26

Changed

• Configured Trusted Publishing for secure automated PyPI deployments

[0.0.4] - 2025-11-26

Changed

• Fixed PyPI deployment issues from v0.0.3

[0.0.3] - 2025-11-25

Changed

• Simplified CI/CD workflows - removed failing tests and strict linting
• Combined release and PyPI publishing into single workflow
• Simplified security scanning to weekly pip-audit only
• Streamlined GitHub Actions configuration

Added

• Comprehensive issue templates (Bug, Feature, Documentation, Support, Grant/Partnership)
• Updated pull request template with clear guidelines
• Community support documentation (SUPPORT.md)
• Funding and sponsorship configuration (FUNDING.yml)
• GitHub configuration README for maintainers
• 10+ new domain-specific cookbook examples (Finance, Healthcare, Cybersecurity, etc.)

Removed

• Redundant scripts folder (8 shell/PowerShell scripts)
• Unnecessary automation workflows (label-issues, mark-answered)
• Excessive issue templates

[0.0.2] - 2025-11-25

Changed

• Updated README with streamlined content and better examples
• Added more notebooks to cookbook
• Improved documentation structure

[0.0.1] - 2024-01-XX

Added

• Core framework architecture
• Universal data ingestion (multiple file formats)
• Semantic intelligence engine (NER, relation extraction, event detection)
• Knowledge graph construction with entity resolution
• 6-stage ontology generation pipeline
• GraphRAG engine for hybrid retrieval
• Multi-agent system infrastructure
• Production-ready quality assurance modules
• Comprehensive documentation with MkDocs
• Cookbook with interactive tutorials
• Support for multiple vector stores (Weaviate, Qdrant, FAISS)
• Support for multiple graph databases (Neo4j, NetworkX, RDFLib)
• Temporal knowledge graph support
• Conflict detection and resolution
• Deduplication and entity merging
• Schema template enforcement
• Seed data management
• Multi-format export (RDF, JSON-LD, CSV, GraphML)
• Visualization tools
• Pipeline orchestration
• Streaming support (Kafka, RabbitMQ, Kinesis)
• Context engineering for AI agents
• Reasoning and inference engine

Documentation

• Getting started guide
• API reference for all modules
• Concepts and architecture documentation
• Use case examples
• Cookbook tutorials
• Community projects showcase

---

Types of Changes

• Added for new features
• Changed for changes in existing functionality
• Deprecated for soon-to-be removed features
• Removed for now removed features
• Fixed for any bug fixes
• Security for vulnerability fixes

Migration Guides

When breaking changes are introduced, migration guides will be provided in the release notes and documentation.

---

For detailed release notes, see GitHub Releases.

Legacy Changelog Snapshot B (Preserved Merge Artifact)

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

• OWLGenerator user-facing schema compatibility fixes (Issue #446):

  • Fixed OWL class/property IRI identifier fallback order to prefer label and then name.
  • Fixed datatype property handling to accept scalar and list range values in rdflib path (including xsd:*, full IRIs, and local names), preventing list-based .startswith() crashes.
  • Fixed generated class/property/domain/range IRIs to use the current ontology dict uri namespace for each generation call (instead of drifting to default namespace manager base URI when per-entity uri is omitted).
  • Fixed subClassOf / subclassOf parent resolution so local class names are expanded to ontology IRIs consistently with domain/range behavior.
  • Added/expanded regression coverage in ests/ontology/test_ontology_comprehensive.py ( est_owl_generator_user_facing_schema_compatibility) for label-first fallback, lowercase subclassOf, datatype range lists, and ontology namespace consistency.

• Fixed: PolicyEngine latest version selection on ContextGraph; AgentContext fallback robustness and secure logging (PR #TBD by @KaifAhmad1)

• Tests: Added ContextGraph fallback and AgentContext smoke tests; full suite passing

• Context Engineering Enhancement (PR #307 by @KaifAhmad1):

  • Comprehensive decision tracking system with full lifecycle management (record → analyze → query → precedent → influence)
  • Advanced KG algorithm integration: centrality analysis, community detection, node embeddings with ContextGraph
  • Enhanced AgentContext with granular feature flags for decision tracking, KG algorithms, and vector store features
  • PolicyException model replacing conflicting Exception name for meaningful business domain modeling
  • GraphStore validation preventing runtime failures with explicit capability checking
  • Hybrid search combining semantic, structural, and category similarity with configurable weights
  • Decision influence analysis with centrality measures and causal chain tracking
  • Policy management with versioning, compliance checking, and exception handling
  • Production-ready architecture with audit trails, security, and scalability features
  • 9 critical bug fixes: logging, security, audit trails, API compatibility, Cypher queries, centrality access, validation, naming
  • Comprehensive documentation with usage guides, production examples, and API references
  • 100% test coverage with all validation tests passing (9/9 tests)
  • Enterprise-grade features for financial services, healthcare, legal, and business domains
  • Complete backward compatibility with existing semantica components
  • Performance optimizations: caching, indexing, and efficient graph operations

• Added PgVector Store Support (PR #303 by @Sameer6305, @KaifAhmad1):

  • Native PostgreSQL vector storage using pgvector extension with full integration
  • Multiple distance metrics: cosine, L2/Euclidean, inner product with automatic score normalization
  • Advanced indexing: HNSW and IVFFlat for approximate nearest neighbor search with tunable parameters
  • JSONB metadata storage with flexible filtering capabilities and batch operations
  • Connection pooling support with psycopg3/psycopg2 fallback and efficient resource management
  • Comprehensive VectorStore integration with backend delegation and unified API
  • Idempotent index creation and table management with safe migration support
  • Production-ready security: SQL injection protection with psycopg_sql.SQL() and input validation
  • Performance optimizations: UUID4-based IDs, batch executemany operations, connection pooling
  • Full backward compatibility with existing vector store implementations
  • 36+ comprehensive test cases with Docker integration and dependency skipping
  • Complete documentation with setup guides, examples, and performance tuning
  • CI/CD integration: resolved benchmark compatibility and fixed documentation links

• Improved Vector Store for Decision Tracking (PR #293 by @KaifAhmad1):

  • Comprehensive decision tracking capabilities with hybrid search combining semantic and structural embeddings
  • New DecisionEmbeddingPipeline for generating semantic and structural embeddings with KG algorithm integration
  • HybridSimilarityCalculator with configurable weights (semantic: 0.7, structural: 0.3)
  • DecisionContext high-level interface for decision management with explainable AI features
  • ContextRetriever with hybrid precedent search and multi-hop reasoning
  • User-friendly convenience API: quick_decision(), find_precedents(), explain(), similar_to(), batch_decisions(), filter_decisions()
  • Knowledge Graph algorithm integration: Node2Vec, PathFinder, CommunityDetector, CentralityCalculator, SimilarityCalculator, ConnectivityAnalyzer
  • Explainable AI with path tracing, confidence scoring, and comprehensive decision explanations
  • Performance optimizations: 0.028s per decision processing, 0.031s search performance, ~0.8KB per decision memory usage
  • 100% backward compatibility maintained with existing VectorStore functionality
  • 34+ comprehensive tests covering all functionality including end-to-end scenarios and performance benchmarks
  • Real-world validation examples for banking and insurance domains
  • Documentation with clear imports, examples, and API references

• Improved Graph Algorithms in KG Module (PR #292 by @KaifAhmad1):

  • Complete algorithm suite with 30+ graph algorithms across 7 categories
  • Node Embeddings: Node2Vec, DeepWalk, Word2Vec for structural similarity analysis
  • Similarity Analysis: Cosine, Euclidean, Manhattan, Correlation metrics with batch processing
  • Path Finding: Dijkstra, A*, BFS, K-shortest paths for route and network analysis
  • Link Prediction: Preferential attachment, Jaccard, Adamic-Adar for network completion
  • Centrality Analysis: Degree, Betweenness, Closeness, PageRank for importance ranking
  • Community Detection: Louvain, Leiden, Label propagation for clustering analysis
  • Connectivity Analysis: Components, bridges, density for network robustness
  • Unified provenance tracking system with GraphBuilderWithProvenance and AlgorithmTrackerWithProvenance
  • Complete execution tracking with metadata, timestamps, and reproducibility IDs
  • Comprehensive test coverage with 5 test suites and 40+ test methods
  • Professional documentation overhaul for all modules and reference documentation
  • Enterprise-ready functionality with error handling and NetworkX compatibility
  • Performance optimizations with sparse matrix operations and batch processing
  • Full backward compatibility maintained with gradual migration support

• Improved Security Configuration with Dependabot:

  • Configured bi-weekly security updates with manual review by @KaifAhmad1
  • Implemented automated security scans (Monday & Thursday at 7 AM IST) with Bandit, Safety, Semgrep
  • Added security-critical package grouping (cryptography, requests, urllib3, certifi, pyopenssl)
  • Enterprise-grade security with audit trail, compliance features, and zero auto-merge
  • Optimized IST timezone scheduling (Security scans: 7 AM IST, PRs: 9 AM IST)
  • Aligned with new Dependabot features: open-source proxy support, smart dependency grouping for Snowflake/Arrow/benchmark features, private registry support, semantic commit prefixes, and latest GitHub security best practices

• ResourceScheduler Deadlock Fix and Performance Improvements (PR #299, #301 by @d4ndr4d3, @KaifAhmad1):

  • Fixed critical deadlock in ResourceScheduler by replacing threading.Lock() with threading.RLock()
  • Resolved nested lock acquisition issue in allocate_resources() → allocate_cpu/memory/gpu() calls
  • Added allocation validation with ValidationError when no resources can be allocated
  • Improved performance by moving progress tracking updates outside lock scope
  • Implemented comprehensive resource cleanup on allocation failures to prevent leaks
  • Added complete regression test suite (6 tests) for deadlock prevention and edge cases
  • Improved error handling and documentation for better operator visibility
  • Zero breaking changes, maintains thread safety and backward compatibility

[0.2.7] - 2026-02-09

Added / Changed

• Snowflake Connector for Data Ingestion (PR #276 by @Sameer6305):

  • Native Snowflake connector with multi-authentication (password, OAuth, key-pair, SSO)
  • Table and query ingestion with pagination, schema introspection, batch processing
  • SQL injection prevention via identifier escaping, OAuth token validation
  • Progress tracking integration, context manager support, document export
  • 24 comprehensive unit tests with mocking, complete documentation and examples
  • Added as optional dependency db-snowflake with snowflake-connector-python>=3.0.0

• Apache Arrow Export Support (PR #273 by @Sameer6305):

  • Added Apache Arrow exporter with explicit schemas, entity/relationship export, compression support
  • Integrated with export module and method registry, Pandas/DuckDB compatible
  • 20 unit tests + 1 integration test, complete documentation with examples

• Comprehensive Benchmark Suite with Regression CLI (PR #289 by @ZohaibHassan16, @KaifAhmad1):

  • 137+ benchmarks across all 10 Semantica modules (Input, Core, Storage, Context, QA, Ontology, etc.)
  • Environment-agnostic design with robust mocking system for CI/CD compatibility
  • Statistical regression detection using Z-score analysis with configurable thresholds
  • Automated performance auditing via GitHub Actions workflow
  • Comprehensive documentation suite (benchmarks.md, architecture guides, usage examples)
  • Zero breaking changes, production-ready with ultra-fast text processing (>10,000 ops/s)
  • Added benchmark runner CLI: python benchmarks/benchmark_runner.py

[0.2.6] - 2026-02-03

Added / Changed

• W3C PROV-O Compliant Provenance Tracking (#254, #246):

  • Comprehensive provenance tracking system with W3C PROV-O compliance across all 17 Semantica modules
  • Core Module: ProvenanceManager, W3C PROV-O schemas, storage backends (InMemory, SQLite), SHA-256 integrity verification
  • Module Integrations: Semantic Extract, LLMs (Groq, OpenAI, HuggingFace, LiteLLM), Pipeline, Context, Ingest, Embeddings, Graph/Vector/Triplet stores, Reasoning, Conflicts, Deduplication, Export, Parse, Normalize, Ontology, Visualization
  • Features: Complete lineage tracking (Document → Chunk → Entity → Relationship → Graph), LLM tracking (tokens, costs, latency), source tracking, bridge axioms for domain transformations
  • Compliance Infrastructure: W3C PROV-O, FDA 21 CFR Part 11, SOX, HIPAA, TNFD
  • Testing: 237 tests covering core functionality, all 17 module integrations, edge cases, backward compatibility
  • Design: Opt-in with provenance=False by default, zero breaking changes, no new dependencies
  • Contributed by @KaifAhmad1

• Enhanced Change Management Module (#248, #243):

  • Enterprise-grade version control for knowledge graphs and ontologies with persistent storage and audit trails
  • Core Classes: TemporalVersionManager (KG versioning), OntologyVersionManager (ontology versioning), ChangeLogEntry (metadata)
  • Storage: SQLite (persistent) and in-memory backends with thread-safe operations
  • Features: SHA-256 checksums, detailed entity/relationship diffs, structural ontology comparison, email validation
  • Compliance Infrastructure: HIPAA, SOX, FDA 21 CFR Part 11 with immutable audit trails
  • Testing: 104 tests (100% pass) - unit, integration, compliance, performance, edge cases
  • Performance: 17.6ms for 10k entities, 510+ ops/sec concurrent, handles 5k+ entity graphs
  • Migration: Backward compatible, simplified class names, zero external dependencies
  • Contributed by @KaifAhmad1

• CSV Ingestion Enhancements (PR #244 by @saloni0318)

  • Auto-detect CSV encoding (chardet) and delimiter (csv.Sniffer)
  • Tolerant decoding and malformed-row handling (on_bad_lines='warn')
  • Optional chunked reading for large files; metadata tracks detected values
  • Expanded unit tests covering delimiters, quoted/multiline fields, header overrides, chunks, and NaN preservation

• Tests: Comprehensive units for TextNormalizer (PR #242 by @ZohaibHassan16)

  • Added focused test coverage for TextNormalizer behavior across inputs

• Tests: Register integration mark and tidy ingest test warnings (PR #241 by @KaifAhmad1)

  • Introduced integration test marker and reduced noisy warnings in ingest tests

• Ingest Unit Tests (#239, #232):

  • Comprehensive unit tests for ingestion modules (file, web, and feed ingestors)
  • Coverage: File scanning (local/cloud S3/GCS/Azure), web ingestion (URL/sitemap/robots.txt), RSS/Atom feed parsing
  • Testing: 998 lines of test code with mocked external dependencies for fast, isolated execution
  • Results: file_ingestor (86%), web_ingestor (86%), feed_ingestor (80%) coverage
  • Covers happy paths, edge cases, and error handling
  • Contributed by @Mohammed2372

Fixed

• Temperature Compatibility Fix (#256, #252):

  • Fixed hardcoded temperature=0.3 that broke compatibility with models requiring specific temperature values (e.g., gpt-5-mini)
  • Added _add_if_set helper method to BaseProvider that only passes parameters when explicitly set
  • When temperature=None, parameter is omitted allowing APIs to use model defaults
  • Updated all 5 providers: OpenAI, Groq, Gemini, Ollama, DeepSeek
  • Reduced code by ~85 lines with cleaner parameter handling
  • Comprehensive test coverage added (10 temperature tests, all passing)
  • Backward compatible - no breaking changes
  • Contributed by @F0rt1s and @IGES-Institut

• JenaStore Empty Graph Bug (#257, #258):

  • Fixed ProcessingError: Graph not initialized when operating on empty (but initialized) graphs
  • Replaced implicit if not self.graph: checks with explicit if self.graph is None: validation in 5 methods (add_triplets, get_triplets, delete_triplet, execute_sparql, serialize)
  • Properly distinguishes None (uninitialized) from empty graphs (initialized with 0 triplets)
  • Unblocks benchmarking suite, fresh deployments, and testing workflows
  • Contributed by @ZohaibHassan16

[0.2.5] - 2026-01-27

Added

• Pinecone Vector Store Support:
  • Implemented native Pinecone support (PineconeStore) with full CRUD capabilities.
  • Added support for serverless and pod-based indexes, namespaces, and metadata filtering.
  • Integrated with VectorStore unified interface and registry.
  • (Closes #219, Resolves #220)
• Configurable LLM Retry Logic:
  • Exposed max_retries parameter in NERExtractor, RelationExtractor, TripletExtractor and low-level extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Defaults to 3 retries to prevent infinite loops during JSON validation failures or API timeouts.
  • Propagated retry configuration through chunked processing helpers to ensure consistent behavior for long documents.
  • Updated 03_Earnings_Call_Analysis.ipynb to use max_retries=3 by default.

Added

• Bring Your Own Model (BYOM) Support:
  • Enabled full support for custom Hugging Face models in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added support for custom tokenizers in HuggingFaceModelLoader to handle models with non-standard tokenization requirements.
  • Implemented robust fallback logic for model selection: runtime options (extract(model=...)) now correctly override configuration defaults.
• Enhanced NER Implementation:
  • Added configurable aggregation strategies (simple, first, average, max) to extract_entities_huggingface for better sub-word token handling.
  • Implemented robust IOB/BILOU parsing to reconstruct entities from raw model outputs when structured output is unavailable.
  • Added confidence scoring for aggregated entities.
• Relation Extraction Improvements:
  • Implemented standard entity marker technique (wrapping subject/object with <subj>, <obj> tags) in extract_relations_huggingface for compatibility with sequence classification models.
  • Added structured output parsing to convert raw model predictions into validated Relation objects.
• Triplet Extraction Completion:
  • Added specialized parsing for Seq2Seq models (e.g., REBEL) in extract_triplets_huggingface to generate structured triplets directly from text.
  • Implemented post-processing logic to clean and validate generated triplets.

Fixed

• LLM Extraction Stability:
  • Fixed infinite retry loops in BaseProvider by strictly enforcing max_retries limit during structured output generation.
  • Resolved stuck execution in earnings call analysis notebooks when using smaller models (e.g., Llama 3 8B) that frequently produce invalid JSON.
• Model Parameter Precedence:
  • Fixed issue where configuration defaults took precedence over runtime arguments in Hugging Face extractors. Runtime options now correctly override config values.
• Import Handling:
  • Fixed circular import issues in test suites by implementing robust mocking strategies.

[0.2.4] - 2026-01-22

Added

• Ontology Ingestion Module:
  • Implemented OntologyIngestor in semantica.ingest for parsing RDF/OWL files (Turtle, RDF/XML, JSON-LD, N3) into standardized OntologyData objects.
  • Added ingest_ontology convenience function and integrated it into the unified ingest(source_type="ontology") interface.
  • Added recursive directory scanning support for batch ontology ingestion.
  • Exposed ingestion tools in semantica.ontology for better discoverability.
  • Added OntologyData dataclass for consistent metadata handling (source path, format, timestamps).
• Documentation:
  • Ontology Usage Guide: Updated ontology_usage.md with comprehensive examples for single-file and directory ingestion.
  • API Reference: Updated ontology.md with OntologyIngestor class documentation and method details.
• Tests:
  • Comprehensive Test Suite: Added tests/ingest/test_ontology_ingestor.py covering all supported formats, error handling, and unified interface integration.
  • Demo Script: Added examples/demo_ontology_ingest.py for end-to-end usage demonstration.

[0.2.3] - 2026-01-20

Fixed

• LLM Relation Extraction Parsing:
  • Fixed relation extraction returning zero relations despite successful API calls to Groq and other providers
  • Normalized typed responses from instructor/OpenAI/Groq to consistent dict format before parsing
  • Added structured JSON fallback when typed generation yields zero relations to avoid silent empty outputs
  • Removed acceptance of extra kwargs (max_tokens, max_entities_prompt) from relation extraction internals
  • Filtered kwargs passed to provider LLM calls to only temperature and verbose
• API Parameter Handling:
  • Limited kwargs forwarded in chunked extraction helper to prevent parameter leakage
  • Ensured minimal, safe parameters are passed to provider calls
• Pipeline Circular Import (Issues #192, #193):
  • Fixed circular import between pipeline_builder and pipeline_validator triggered during semantica.pipeline import
  • Lazy-loaded PipelineValidator inside PipelineBuilder.__init__ and guarded type hints with TYPE_CHECKING
  • Ensured from semantica.deduplication import DuplicateDetector no longer fails even when pipeline module is imported
• JupyterLab Progress Output (Issue #181):
  • Added SEMANTICA_DISABLE_JUPYTER_PROGRESS environment variable to disable rich Jupyter/Colab progress tables
  • When enabled, progress falls back to console-style output, preventing infinite scrolling and JupyterLab out-of-memory errors

Added

• Comprehensive Test Suite:
• • Added unit tests (tests/test_relations_llm.py) with mocked LLM provider covering both typed and structured response paths
• • Added integration tests (tests/integration/test_relations_groq.py) for real Groq API calls with environment variable API key
• • Tests validate relation extraction completion and result parsing across different response formats
• Amazon Neptune Dev Environment:
• • Added CloudFormation template (cookbook/introduction/neptune-setup.yaml) to provision a dev Neptune cluster with public endpoint and IAM auth enabled
• • Documented deployment, cost estimates, and IAM User vs IAM Role best practices in cookbook/introduction/21_Amazon_Neptune_Store.ipynb
• • Added cfn-lint to .pre-commit-config.yaml for validating CloudFormation templates while excluding neptune-setup.yaml from generic YAML linters
• Vector Store High-Performance Ingestion:
• • Added VectorStore.add_documents for high-throughput ingestion with automatic embedding generation, batching, and parallel processing
• • Added VectorStore.embed_batch helper for generating embeddings for lists of texts without immediately storing them
• • Enabled default parallel ingestion in VectorStore with max_workers=6 for common workloads
• • Added dedicated documentation page docs/vector_store_usage.md describing high-performance vector store usage and configuration
• • Added tests/vector_store/test_vector_store_parallel.py covering parallel vs sequential performance, error handling, and edge cases for add_documents and embed_batch

Changed

• Relation Extraction API:
• • Simplified parameter interface by removing unused kwargs that were previously ignored
• • Improved error handling and verbose logging for debugging relation extraction issues
• • Enhanced robustness of post-response parsing across different LLM providers
• Vector Store Defaults and Examples:
• • Standardized VectorStore default concurrency to max_workers=6 for parallel ingestion
• • Updated vector store reference documentation and usage guides to rely on implicit defaults instead of requiring manual max_workers configuration in examples

[0.2.2] - 2026-01-15

Added

• Parallel Extraction Engine:
  • Implemented high-throughput parallel batch processing across all core extractors (NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticNetworkExtractor) using concurrent.futures.ThreadPoolExecutor.
  • Added max_workers configuration parameter (default: 1) to all extractor extract() methods, allowing users to tune concurrency based on available CPU cores or API rate limits.
  • Parallel Chunking: Implemented parallel processing for large document chunking in _extract_entities_chunked and _extract_relations_chunked, significantly reducing latency for long-form text analysis.
  • Thread-Safe Progress Tracking: Enhanced ProgressTracker to handle concurrent updates from multiple threads without race conditions during batch processing.
• Semantic Extract Performance & Regression:
  • Added edge-case regression suite covering max worker defaults, LLM prompt entity filtering, and extractor reuse.
  • Added a runnable real-use-case benchmark script for batch latency across NERExtractor, RelationExtractor, TripletExtractor, EventDetector, SemanticAnalyzer, and SemanticNetworkExtractor.
  • Added Groq LLM smoke tests that exercise LLM-based entities/relations/triplets when GROQ_API_KEY is available via environment configuration.

Security

• Credential Sanitization:
  • Removed hardcoded API keys from 8 cookbook notebooks to prevent secret leakage.
  • Enforced environment variable usage for GROQ_API_KEY across all examples.
• Secure Caching:
  • Updated ExtractionCache to exclude sensitive parameters (e.g., api_key, token, password) from cache key generation, preventing secret leakage and enabling safe cache sharing.
  • Upgraded cache key hashing algorithm from MD5 to SHA-256 for enhanced collision resistance and security.

Changed

• Gemini SDK Migration:
  • Migrated GeminiProvider to use the new google-genai SDK (v0.1.0+) to address deprecation warnings.
  • Implemented graceful fallback to google.generativeai for backward compatibility.
• Dependency Resolution:
  • Pinned opentelemetry-api and opentelemetry-sdk to 1.37.0 to resolve pip conflicts.
  • Updated protobuf and grpcio constraints for better stability.
• Entity Filtering Scope:
  • Removed entity filtering from non-LLM extraction flows to avoid accuracy regressions.
  • Applied entity downselection only to LLM relation prompt construction, while matching returned entities against the full original entity list.
• Batch Concurrency Defaults:
  • Standardized max_workers defaulting across semantic_extract and tuned for low-latency: ML-backed methods default to single-worker, while pattern/regex/rules/LLM/huggingface methods use a higher parallelism default capped by CPU.
  • Raised the global optimization.max_workers default to 8 for better throughput on batch workloads.

Performance

• Bottleneck Optimization (GitHub Issue #186):
  • Resolved Bottleneck #1 (Sequential Processing): Replaced sequential for loops with parallel execution for both document-level batches and intra-document chunks.
  • Performance Gains: Achieved ~1.89x speedup in real-world extraction scenarios (tested with Groq llama-3.3-70b-versatile on standard datasets).
  • Initialization Optimization: Refactored test suite to use class-level setUpClass for LLM provider initialization, eliminating redundant API client creation overhead.
• Low-Latency Entity Matching:
  • Avoided heavyweight embedding stack imports on common matches by improving fast matching heuristics and short-circuiting before embedding similarity.
  • Optimized entity matching to prioritize exact/substring/word-boundary matches and only fall back to embedding similarity when needed, reducing CPU overhead in LLM relation/triplet mapping.

[0.2.1] - 2026-01-12

Fixed

• LLM Output Stability (Bug #176):
  • Fixed incomplete JSON output issues by correctly propagating max_tokens parameter in extract_relations_llm.
  • Implemented automatic error handling that halves chunk sizes and retries when LLM context or output limits are exceeded.
  • Fixed AttributeError in provider integration by ensuring consistent parameter passing via **kwargs.
• Constraint Relaxations:
  • Removed hardcoded max_length constraints from Entity, Relation, and Triplet classes to support long-form semantic extraction (e.g., long descriptions or names).
• Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
• Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
• Fixed dependency compatibility issues by pinning protobuf>=5.29.1,<7.0 and grpcio>=1.71.2.
• Added missing dependencies GitPython and chardet to pyproject.toml.
• Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Changed

• Chunking Defaults:
  • Increased default max_text_length for auto-chunking to 64,000 characters (from 32k/16k) for OpenAI, Anthropic, Gemini, Groq, and DeepSeek providers.
  • Unified chunking logic across extract_entities_llm, extract_relations_llm, and extract_triplets_llm.
• Groq Support:
  • Standardized Groq provider defaults to use llama-3.3-70b-versatile with a 64k context window.
  • Added native support for max_tokens and max_completion_tokens to prevent output truncation.

Added

• Testing:
  • Added tests/reproduce_issue_176.py to validate max_tokens propagation and chunking behavior across all extractors.

[0.2.0] - 2026-01-10

Added

• Amazon Neptune Support:
  • Added AmazonNeptuneStore providing Amazon Neptune graph database integration via Bolt protocol and OpenCypher.
  • Implemented NeptuneAuthTokenManager extending Neo4j AuthManager for AWS IAM SigV4 signing with automatic token refresh.
  • Added robust connection handling: retry logic with backoff for transient errors (signature expired, connection closed) and driver recreation.
  • Added graph-amazon-neptune optional dependency group (boto3, neo4j).
  • Comprehensive test suite covering all GraphStore interface methods.
• Docling Integration:
  • Added DoclingParser in semantica.parse for high-fidelity document parsing using the Docling library.
  • Supports multi-format parsing (PDF, DOCX, PPTX, XLSX, HTML, images) with superior table extraction and structure understanding.
  • Implemented as a standalone parser supporting local execution, OCR, and multiple export formats (Markdown, HTML, JSON).
• Robust Extraction Fallbacks:
  • Implemented comprehensive fallback chains ("ML/LLM" -> "Pattern" -> "Last Resort") across NERExtractor, RelationExtractor, and TripletExtractor to prevent empty result lists.
  • Added "Last Resort" pattern matching in NERExtractor to identify capitalized words as generic entities when all other methods fail.
  • Added "Last Resort" adjacency-based relation extraction in RelationExtractor to create weak connections between adjacent entities if no relations are found.
  • Added fallback logic in TripletExtractor to convert relations to triplets or use rule-based extraction if standard methods fail.
• Provenance & Tracking:
  • Added count tracking to batch processing logs in NERExtractor, RelationExtractor, and TripletExtractor.
  • Added batch_index and document_id to the metadata of all extracted entities, relations, triplets, semantic roles, and clusters for better traceability.
• Semantic Extract Improvements:
  • Introduced auto-chunking for long text processing in LLM extraction methods (extract_entities_llm, extract_relations_llm, extract_triplets_llm).
  • Added silent_fail parameter to LLM extraction methods for configurable error handling.
  • Implemented robust JSON parsing and automatic retry logic (3 attempts with exponential backoff) in BaseProvider for all LLM providers.
  • Enhanced GroqProvider with better diagnostics and connectivity testing.
  • Added comprehensive entity, relation, and triplet deduplication for chunked extraction.
  • Added semantica/semantic_extract/schemas.py with canonical Pydantic models for consistent structured output.
• Testing:
  • Added comprehensive robustness test suite tests/semantic_extract/test_robustness_fallback.py for validating extraction fallbacks and metadata propagation.
  • Added comprehensive unit test suite tests/embeddings/test_model_switching.py for verifying dynamic model transitions and dimension updates.
  • Added end-to-end integration test suite for Knowledge Graph pipeline validation (GraphBuilder -> EntityResolver -> GraphAnalyzer).
• Other:
  • Added missing dependencies GitPython and chardet to pyproject.toml.
  • Robustified ID extraction across CentralityCalculator, CommunityDetector, and ConnectivityAnalyzer to handle various entity formats.
  • Improved Entity class hashability and equality logic in utils/types.py.

Changed

• Deduplication & Conflict Logic:
  • Removed internal deduplication logic from NERExtractor, RelationExtractor, and TripletExtractor.
  • Removed consistency/conflict checking from ExtractionValidator to defer to dedicated semantica/conflicts module.
  • Removed _deduplicate_* methods from semantica/semantic_extract/methods.py.
• Batch Processing & Consistency:
  • Standardized batch processing across all extractors (NERExtractor, RelationExtractor, TripletExtractor, SemanticNetworkExtractor, EventDetector, SemanticAnalyzer, CoreferenceResolver) using a unified extract/analyze/resolve method pattern with progress tracking.
  • Added provenance metadata (batch_index, document_id) to SemanticNetwork nodes/edges, Event objects, SemanticRole results, CoreferenceChain mentions, and SemanticCluster (tracking source document_ids).
  • Updated SemanticClusterer.cluster and SemanticAnalyzer.cluster_semantically to accept list of dictionaries (with content and id keys) for better document tracking during clustering.
  • Removed legacy check_triplet_consistency from TripletExtractor.
  • Removed validate_consistency and _check_consistency from ExtractionValidator.
• Weighted Scoring:
  • Clarified weighted confidence scoring (50% Method Confidence + 50% Type Similarity) in comments.
  • Explicitly labeled "Type Similarity" as "user-provided" in code comments to remove ambiguity.
• Refactoring:
  • Fixed orchestrator lazy property initialization and configuration normalization logic in Orchestrator.
  • Verified and aligned FileObject.text property usage in GraphRAG notebooks for consistent content decoding.

Fixed

• Critical Fixes:
  • Resolved NameError in extraction_validator.py by adding missing Union import.
  • Resolved issues where extractors would return empty lists for valid input text when primary extraction methods failed.
  • Fixed metadata initialization issue in batch processing where batch_index and document_id were occasionally missing from extracted items.
  • Ensured LLMExtraction methods (enhance_entities, enhance_relations) return original input instead of failing or returning empty results when LLM providers are unavailable.
• Component Fixes:
  • Fixed model switching bug in TextEmbedder where internal state was not cleared, preventing dynamic updates between fastembed and sentence_transformers (#160).
  • Implemented model-intrinsic embedding dimension detection in TextEmbedder to ensure consistency between models and vector databases.
  • Updated set_model to properly refresh configuration and dimensions during model switches.
  • Fixed TypeError: unhashable type: 'Entity' in GraphAnalyzer when processing graphs with raw Entity objects or dictionaries in relationships (#159).
  • Resolved AssertionError in orchestrator tests by aligning test mocks with production component usage.
  • Fixed dependency compatibility issues by pinning protobuf==4.25.3 and grpcio==1.67.1.
  • Fixed a bug in TripletExtractor where the validate_triplets method was shadowed by an internal attribute.
  • Fixed incorrect TextSplitter import path in the semantic_extract.methods module.

[0.1.1] - 2026-01-05

Added

• Exported DoclingParser and DoclingMetadata from semantica.parse for easier access.
• Added comprehensive DoclingParser usage examples to README and documentation.
• Added Windows-specific troubleshooting note for PyTorch DLL issues.

Fixed

• Fixed DoclingParser import/export issues across platforms (Windows, Linux, Google Colab).
• Improved error messaging when optional docling dependency is missing.
• Fixed versioning inconsistencies across the framework.

[0.1.0] - 2025-12-31

Added

• New command-line interface (semantica CLI) with support for knowledge base building and info commands.
• Integrated FastAPI-based REST API server for remote access to framework functionality.
• Dedicated background worker component for scalable task processing and pipeline execution.
• Framework-level versioning configuration for PyPI distribution.
• Automated release workflow with Trusted Publishing support.

Changed

• Updated versioning across the framework to 0.1.0.
• Refined entry point configurations in pyproject.toml.
• Improved lazy module loading for core framework components.

[0.0.5] - 2025-11-26

Changed

• Configured Trusted Publishing for secure automated PyPI deployments

[0.0.4] - 2025-11-26

Changed

• Fixed PyPI deployment issues from v0.0.3

[0.0.3] - 2025-11-25

Changed

• Simplified CI/CD workflows - removed failing tests and strict linting
• Combined release and PyPI publishing into single workflow
• Simplified security scanning to weekly pip-audit only
• Streamlined GitHub Actions configuration

Added

• Comprehensive issue templates (Bug, Feature, Documentation, Support, Grant/Partnership)
• Updated pull request template with clear guidelines
• Community support documentation (SUPPORT.md)
• Funding and sponsorship configuration (FUNDING.yml)
• GitHub configuration README for maintainers
• 10+ new domain-specific cookbook examples (Finance, Healthcare, Cybersecurity, etc.)

Removed

• Redundant scripts folder (8 shell/PowerShell scripts)
• Unnecessary automation workflows (label-issues, mark-answered)
• Excessive issue templates

[0.0.2] - 2025-11-25

Changed

• Updated README with streamlined content and better examples
• Added more notebooks to cookbook
• Improved documentation structure

[0.0.1] - 2024-01-XX

Added

• Core framework architecture
• Universal data ingestion (multiple file formats)
• Semantic intelligence engine (NER, relation extraction, event detection)
• Knowledge graph construction with entity resolution
• 6-stage ontology generation pipeline
• GraphRAG engine for hybrid retrieval
• Multi-agent system infrastructure
• Production-ready quality assurance modules
• Comprehensive documentation with MkDocs
• Cookbook with interactive tutorials
• Support for multiple vector stores (Weaviate, Qdrant, FAISS)
• Support for multiple graph databases (Neo4j, NetworkX, RDFLib)
• Temporal knowledge graph support
• Conflict detection and resolution
• Deduplication and entity merging
• Schema template enforcement
• Seed data management
• Multi-format export (RDF, JSON-LD, CSV, GraphML)
• Visualization tools
• Pipeline orchestration
• Streaming support (Kafka, RabbitMQ, Kinesis)
• Context engineering for AI agents
• Reasoning and inference engine

Documentation

• Getting started guide
• API reference for all modules
• Concepts and architecture documentation
• Use case examples
• Cookbook tutorials
• Community projects showcase

---

Types of Changes

• Added for new features
• Changed for changes in existing functionality
• Deprecated for soon-to-be removed features
• Removed for now removed features
• Fixed for any bug fixes
• Security for vulnerability fixes

Migration Guides

When breaking changes are introduced, migration guides will be provided in the release notes and documentation.

---

For detailed release notes, see GitHub Releases.