mengzhx/semantica
1
0
Fork 0
mirror of https://github.com/semantica-agi/semantica.git synced 2026-08-29 04:26:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs: add choose-your-module onboarding guide (#651)

Browse Source 
* docs: add choose-your-module onboarding guide

* fix(docs): correct export code examples against actual API signatures

- export_to_rdf() returns a string; use export() for file output
- format="json-ld" is invalid; correct value is "jsonld"
- ParquetExporter/LPGExporter/ArangoAQLExporter take file_path as a
  required positional arg, not output= / output_dir= kwargs
- ArangoAQLExporter().export(graph) was missing file_path entirely,
  which would raise TypeError at runtime
- Remove misleading 'with provenance embedded' comment (no such param)

---------

Co-authored-by: KaifAhmad1 <kaifahmad087@gmail.com>
This commit is contained in:
Sameer Kadam
2026-06-19 13:15:06 +05:30
committed by GitHub
co-authored by KaifAhmad1
parent 12d61b92df
commit 926d4c1653
4 changed files with 338 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/choose-your-module.md
+327
View File
@@ -0,0 +1,327 @@
---
title: "Choose the Right Module"
description: "Map your goal to the right Semantica module in under 30 seconds."
icon: "compass"
---
<Info>
Every module works independently — import only what you need. This page maps developer goals to starting points. The [Module Reference](modules) covers every module in depth.
</Info>
## Quick Reference
Find your goal below. The **Module** column is your import path; **Key class** is what you instantiate first.
| I want to... | Module | Key class |
| :------------ | :------ | :--------- |
| Load a PDF, DOCX, HTML, CSV, or archive | `ingest` | `FileIngestor` |
| Crawl a website | `ingest` | `WebIngestor` |
| Load Parquet files or partitioned datasets | `ingest` | `ParquetIngestor` |
| Ingest XML with schema validation | `ingest` | `XMLIngestor` |
| Ingest from SQL, Snowflake, Kafka, or email | `ingest` | `DBIngestor`, `SnowflakeIngestor`, `StreamIngestor` |
| Extract clean text and tables from a document | `parse` | `DocumentParser` |
| Parse complex PDFs with OCR or multi-column layout | `parse` | `DoclingParser` |
| Chunk text for embedding or RAG | `split` | `TextSplitter` |
| Normalize text, dates, entities, or encodings | `normalize` | `TextNormalizer`, `EntityNormalizer` |
| Find named entities (people, orgs, locations) in text | `semantic_extract` | `NERExtractor` |
| Extract typed relationships from text | `semantic_extract` | `RelationExtractor` |
| Extract RDF subject–predicate–object triplets | `semantic_extract` | `TripletExtractor` |
| Build a queryable knowledge graph | `kg` | `GraphBuilder` |
| Add time-validity (`valid_from` / `valid_until`) to facts | `kg` | `TemporalGraphQuery` |
| Run graph algorithms (centrality, communities, paths) | `kg` | `GraphAnalyzer`, `CentralityCalculator` |
| Generate vector embeddings | `embeddings` | `EmbeddingGenerator` |
| Store and search vectors | `vector_store` | `VectorStore` |
| Persist a graph in Neo4j or FalkorDB | `graph_store` | `Neo4jStore`, `FalkorDBStore` |
| Store RDF triples and query with SPARQL | `triplet_store` | `TripletStore` |
| Deduplicate entities across sources | `deduplication` | `DuplicateDetector`, `EntityMerger` |
| Detect and resolve contradictory facts | `conflicts` | `ConflictDetector`, `ConflictResolver` |
| Give an AI agent persistent memory | `context` | `AgentContext` |
| Ground LLM responses in a knowledge graph (GraphRAG) | `context` | `AgentContext.query_with_reasoning()` |
| Record AI decisions with a full audit trail | `context` | `AgentContext.record_decision()` |
| Search past decisions before making a new one | `context` | `AgentContext.find_precedents()` |
| Trace the causal chain of a decision | `context` | `AgentContext.get_causal_chain()` |
| Track where every fact came from (W3C PROV-O) | `provenance` | `ProvenanceManager` |
| Version a graph with checksums and rollback | `change_management` | `TemporalVersionManager` |
| Auto-generate an OWL schema from a graph | `ontology` | `OntologyGenerator` |
| Validate a graph against SHACL constraints | `ontology` | `SHACLGenerator`, `OntologyValidator` |
| Derive new facts from existing knowledge | `reasoning` | `Reasoner`, `GraphReasoner` |
| Export to RDF Turtle, JSON-LD, or N-Triples | `export` | `RDFExporter` |
| Export to Parquet for Spark / BigQuery | `export` | `ParquetExporter` |
| Export for ArangoDB | `export` | `ArangoAQLExporter` |
| Export to Neo4j or Memgraph via Cypher | `export` | `LPGExporter` |
| Visualize a knowledge graph interactively | `visualization` | `KGVisualizer` |
| Run a reproducible multi-step pipeline | `pipeline` | `PipelineBuilder` |
| Use Semantica from Claude Desktop or Cursor | `mcp_server` | `semantica-mcp` |
| Bootstrap a graph from verified seed data | `seed` | `SeedDataManager` |
| Extend Semantica with a custom component | `core` | `PluginRegistry` |
## Goal-by-Goal Starting Points
Pick your goal to see the minimum imports and a working skeleton.
<Tabs>
<Tab title="Build a Knowledge Graph">
Turn documents, web pages, or databases into a structured, queryable graph.
**Pipeline:** `ingest` → `parse` → `semantic_extract` → `kg`
```python
from semantica.ingest import FileIngestor
from semantica.parse import DocumentParser
from semantica.semantic_extract import NERExtractor, RelationExtractor
from semantica.kg import GraphBuilder
sources = FileIngestor().ingest("report.pdf")
parsed = DocumentParser().parse_document("report.pdf")
# No API key required — pattern-based extraction
entities = NERExtractor(method="pattern").extract(parsed)
relationships = RelationExtractor(method="rule").extract(parsed, entities=entities)
graph = GraphBuilder(merge_entities=True).build(
sources=[{"entities": entities, "relationships": relationships}]
)
print(f"{len(graph.nodes)} nodes, {len(graph.edges)} edges")
```
<Tip>
Pass `method="pattern"` to `NERExtractor` for zero-cost, zero-API-key extraction. Switch to `method="llm"` with any of the supported providers for higher recall.
</Tip>
**Next:** [Quickstart →](quickstart) — full pipeline with visualization and export.
</Tab>
<Tab title="Build GraphRAG">
Ground every LLM response in a structured knowledge graph. Every claim links back to a source node.
**Module:** `context`
```python
from semantica.context import AgentContext, ContextGraph
from semantica.vector_store import VectorStore
from semantica.llms import Groq
llm = Groq(model="llama-3.3-70b-versatile")
context = AgentContext(
vector_store=VectorStore(backend="faiss", dimension=768),
knowledge_graph=ContextGraph(advanced_analytics=True),
)
# Store facts — retrieval uses both vectors and graph structure
context.store("Apple Inc. was co-founded by Steve Jobs in 1976 in Cupertino.")
# GraphRAG query with multi-hop reasoning trace
result = context.query_with_reasoning(
"Who co-founded Apple?",
llm_provider=llm,
max_hops=2,
)
print(result["response"]) # grounded answer
print(result["reasoning_path"]) # multi-hop trace
```
**Next:** [Context module reference →](reference/context)
</Tab>
<Tab title="Add Agent Memory">
Give an AI agent persistent memory, decision tracking, and precedent search across sessions.
**Module:** `context`
```python
from semantica.context import AgentContext, ContextGraph
from semantica.vector_store import VectorStore
context = AgentContext(
vector_store=VectorStore(backend="faiss", dimension=768),
knowledge_graph=ContextGraph(advanced_analytics=True),
decision_tracking=True, # required to use record_decision()
)
# Store a memory
context.store("GPT-4 outperforms GPT-3.5 on reasoning benchmarks by 40%.")
# Record a decision with full causal context
decision_id = context.record_decision(
category="model_selection",
scenario="Choose LLM for production reasoning pipeline",
reasoning="GPT-4 benchmark advantage justifies cost increase",
outcome="selected_gpt4",
confidence=0.91,
)
# Search past decisions before making a new one
precedents = context.find_precedents("model selection", limit=5)
# Trace what happened downstream from this decision
chain = context.get_causal_chain(decision_id, direction="downstream")
```
<Note>
`decision_tracking=True` is required. Without it, `record_decision()` raises `RuntimeError`.
</Note>
**Next:** [Context module reference →](reference/context)
</Tab>
<Tab title="Track Provenance">
W3C PROV-O lineage on every fact: source document, extraction method, timestamp, and checksum.
**Modules:** `provenance`, `change_management`
```python
from semantica.provenance import ProvenanceManager
prov = ProvenanceManager()
# Track an entity with full source details
prov.track_entity(
entity_id="entity_1",
source="DOI:10.1371/journal.pone.0023601",
source_location="Figure 2",
confidence=0.92,
)
# Retrieve the complete lineage for this entity
lineage = prov.get_lineage("entity_1")
# Version-control the graph with SHA-256 checksums
from semantica.change_management import TemporalVersionManager
manager = TemporalVersionManager()
snapshot = manager.create_snapshot(kg, "v1.0", "user@example.com", "Initial build")
diff = manager.diff("v1.0", "v1.1")
```
**Next:** [Provenance reference →](reference/provenance) · [Change Management reference →](reference/change_management)
</Tab>
<Tab title="Export">
Serialize your knowledge graph for the semantic web, analytics platforms, or graph databases.
**Module:** `export`
```python
from semantica.export import RDFExporter, ParquetExporter, LPGExporter, ArangoAQLExporter
# RDF — multiple serialization formats
RDFExporter().export(graph, "graph.ttl", format="turtle")
RDFExporter().export(graph, "graph.jsonld", format="jsonld")
# Parquet — for Spark, BigQuery, Databricks, Snowflake
ParquetExporter().export(graph, "output/graph.parquet")
# Neo4j / Memgraph via Cypher
LPGExporter().export(graph, "graph.cypher")
# ArangoDB AQL inserts
ArangoAQLExporter().export(graph, "graph.aql")
```
**Formats:** Turtle · JSON-LD · N-Triples · RDF/XML · Parquet · Cypher · Arrow · OWL · CSV · ArangoDB AQL
**Next:** [Export module reference →](reference/export)
</Tab>
<Tab title="MCP — Claude / Cursor">
Use Semantica from Claude Desktop, Cursor, VS Code, or any MCP-aware tool — no Python code required after setup. 12 tools available instantly.
**Step 1 — Install:**
```bash
pip install semantica
```
**Step 2 — Add to your MCP client config:**
<CodeGroup>
```json Claude Desktop / Windsurf / Cline
{
"mcpServers": {
"semantica": {
"command": "semantica-mcp"
}
}
}
```
```json Cursor / VS Code / Continue
{
"mcpServers": {
"semantica": {
"command": "semantica-mcp",
"env": {
"SEMANTICA_KG_PATH": "/path/to/my_graph.json"
}
}
}
}
```
</CodeGroup>
**Available tools:** `extract_entities` · `extract_relations` · `add_entity` · `add_relationship` · `record_decision` · `query_decisions` · `find_precedents` · `get_causal_chain` · `run_reasoning` · `get_graph_analytics` · `export_graph` · `get_graph_summary`
<Warning>
Set `SEMANTICA_KG_PATH` to persist your graph across restarts. Without it, all data is lost when the server process exits.
</Warning>
**Next:** [MCP Server reference →](reference/mcp_server)
</Tab>
</Tabs>
## Still Unsure?
<AccordionGroup>
<Accordion title="Knowledge graph vs. vector store — which do I need?" icon="scale-balanced">
Use a **knowledge graph** (`kg`) when you need structured reasoning, multi-hop traversal, provenance, or compliance audit trails.
Use a **vector store** (`vector_store`) when you need fast fuzzy similarity search over large text corpora and relationships between items don't matter.
Use **both together** via `AgentContext` (GraphRAG) to get grounded LLM responses where every claim traces back to a source node.
See also: [Core Concepts](concepts)
</Accordion>
<Accordion title="I just want to run something quickly." icon="rocket">
Start with the [Quickstart](quickstart). It builds a complete pipeline (ingest → parse → extract → graph → visualize → export) with no API key required.
</Accordion>
<Accordion title="I'm adding Semantica to an existing agent — what's the minimum?" icon="plug">
Add `AgentContext`. It wraps your existing agent with memory, decision tracking, and precedent search — no changes to your LLM provider or agent framework needed.
```python
from semantica.context import AgentContext, ContextGraph
from semantica.vector_store import VectorStore
context = AgentContext(
vector_store=VectorStore(backend="faiss", dimension=768),
knowledge_graph=ContextGraph(advanced_analytics=True),
decision_tracking=True,
)
```
[Context module reference →](reference/context)
</Accordion>
<Accordion title="I need a compliance-ready pipeline — what's the minimum stack?" icon="shield-check">
| Layer | Module | Key class |
| :---- | :------ | :--------- |
| Ingestion | `ingest` | `FileIngestor` |
| Extraction | `semantic_extract` | `NERExtractor` |
| Graph | `kg` | `GraphBuilder` |
| Lineage | `provenance` | `ProvenanceManager` |
| Versioning | `change_management` | `TemporalVersionManager` |
| Audit export | `export` | `RDFExporter` |
Supports HIPAA, SOX, GDPR, and FDA 21 CFR Part 11 audit requirements.
</Accordion>
</AccordionGroup>
---
- [Quickstart](quickstart) — Full pipeline in 5 minutes.
- [Module Reference](modules) — Every module with examples and common chains.
- [API Reference](reference/context) — Complete class and method documentation.
docs/docs.json
+5 -3
View File
@@ -65,6 +65,7 @@
"pages": [
"concepts",
"modules",
"choose-your-module",
"glossary"
]
},
@@ -104,10 +105,11 @@
"group": "Get Started",
"pages": [
"installation",
"cli-setup",
"explorer-setup",
"getting-started",
"quickstart",
"getting-started"
"choose-your-module",
"cli-setup",
"explorer-setup"
]
}
]
docs/getting-started.md
+2 -15
View File
@@ -202,22 +202,9 @@ Semantica uses a modular, layered architecture: import only what you need.
- **[Output Layer](reference/export)** — Deliver results downstream. Modules: `export`, `visualization`, `pipeline`, `explorer`
## "Which module do I need?" Quick Reference
## Which Module Do I Need?
| I want to... | Module | Key class |
| :------------ | :------ | :--------- |
| Load a PDF / web page / database | `ingest` | `FileIngestor`, `WebIngestor` |
| Extract text and tables from a PDF | `parse` | `DocumentParser`, `DoclingParser` |
| Find entities in text | `semantic_extract` | `NERExtractor` |
| Build a knowledge graph | `kg` | `GraphBuilder` |
| Store and search vectors | `vector_store` | `VectorStore` |
| Give my agent persistent memory | `context` | `AgentContext` |
| Record AI decisions with audit trail | `context` | `AgentContext.record_decision()` |
| Query my graph with natural language | `reasoning` | `GraphReasoner` |
| Export to RDF / Neo4j / Parquet | `export` | `RDFExporter`, `LPGExporter` |
| Visualize a knowledge graph | `visualization` | `KGVisualizer` |
| Run a reproducible pipeline | `pipeline` | `PipelineBuilder` |
| Use Semantica from Claude Desktop | `mcp_server` | `semantica-mcp` |
See the [Choose the Right Module](choose-your-module) guide — it maps 35+ developer goals to the right starting point across all 27 modules, with working code for the most common paths.
## Next Steps
docs/modules.md
+4
View File
@@ -8,6 +8,10 @@ icon: "puzzle-piece"
Looking for a quick reference? Jump to the [Module Index](#module-index) at the bottom.
</Info>
<Tip>
Not sure which module to use? The [Choose the Right Module](choose-your-module) guide maps 35+ developer goals to modules with code examples — start there if you're orienting for the first time.
</Tip>
Semantica is organized into **27 modules** across six logical layers. Each module is independently importable: you never pay for what you don't use.
## Architecture Overview