Knowledge Capture

Overview

Knowledge Capture component helps you capture and structure knowledge from unstructured data sources. It provides intelligent organization and structuring capabilities for raw information.

What is Knowledge Capture?

Knowledge Capture is the process of extracting meaningful information from unstructured sources and organizing it into a structured format that can be easily queried and analyzed.

Key Features

Intelligent Structuring: Automatically organizes unstructured data
Entity Recognition: Identifies and extracts key entities
Relationship Mapping: Maps relationships between entities
Context Preservation: Maintains context during capture
Multiple Input Formats: Supports various data sources

Basic Usage

import { KnowledgeCapture } from 'gaik-toolkit';

// Create a knowledge capture instance
const capture = new KnowledgeCapture();

// Capture knowledge from text
const knowledge = await capture.captureFromText(text);

console.log(knowledge.entities);
console.log(knowledge.relationships);

API Reference

`KnowledgeCapture`

Main class for capturing and structuring knowledge.

Constructor

const capture = new KnowledgeCapture(options);

Options:

language (string): Language for processing (default: 'en')
entityTypes (array): Types of entities to extract
includeRelationships (boolean): Include relationship mapping (default: true)

Methods

`captureFromText(text: string): Promise<Knowledge>`

Captures knowledge from text content.

Parameters:

text: Text content to process

Returns: Promise resolving to Knowledge object

Example:

const knowledge = await capture.captureFromText('Sample text...');

`captureFromDocument(document: ParseResult): Promise<Knowledge>`

Captures knowledge from a parsed document.

Parameters:

document: ParseResult from Parser component

Returns: Promise resolving to Knowledge object

`Knowledge`

Result object containing captured knowledge.

Properties:

entities (array): Extracted entities with metadata
relationships (array): Relationships between entities
concepts (array): Key concepts identified
summary (string): Generated summary
metadata (object): Additional metadata

Usage Examples

Capturing from Parsed Documents

import { Parser, KnowledgeCapture } from 'gaik-toolkit';

const parser = new Parser();
const capture = new KnowledgeCapture();

// Parse document
const document = await parser.parse('document.pdf');

// Capture knowledge
const knowledge = await capture.captureFromDocument(document);

console.log('Entities:', knowledge.entities);
console.log('Concepts:', knowledge.concepts);

Custom Entity Types

const capture = new KnowledgeCapture({
  entityTypes: ['person', 'organization', 'location', 'date'],
  includeRelationships: true
});

const knowledge = await capture.captureFromText(text);

Working with Relationships

const knowledge = await capture.captureFromText(text);

// Access relationships
knowledge.relationships.forEach(rel => {
  console.log(`${rel.source} -> ${rel.type} -> ${rel.target}`);
});

Advanced Features

Entity Filtering

const knowledge = await capture.captureFromText(text);

// Filter entities by type
const people = knowledge.entities.filter(e => e.type === 'person');
const organizations = knowledge.entities.filter(e => e.type === 'organization');

Knowledge Graph

// Build a knowledge graph
const graph = capture.buildGraph(knowledge);

// Query the graph
const connections = graph.findConnections(entity1, entity2);

Integration with Other Components

Knowledge Capture works seamlessly with other GAIK toolkit components:

import { Parser, KnowledgeCapture, KnowledgeExtraction } from 'gaik-toolkit';

// Parse -> Capture -> Extract workflow
const document = await parser.parse('document.pdf');
const knowledge = await capture.captureFromDocument(document);
const insights = await extractor.extract(knowledge, query);

Next Steps

Explore Knowledge Extraction
Check out Semantic Video Search Example
Learn about the Parser component

On this page