A professional-grade pipeline for extracting entities, relations, and events from raw text into normalized, graph-ready structured formats.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install information-extraction
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install information-extraction using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Information Extraction?
The Information Extraction skill is a specialized tool within the Openclaw Skills ecosystem designed to bridge the gap between unstructured prose and structured intelligence. It enables AI agents to parse plain text or Markdown and identify specific entities, their attributes, and the complex relationships connecting them. By utilizing a semi-automatic pipeline, it ensures that raw data from reports, transcripts, and web content is converted into a high-fidelity intermediate structure before final export.
This skill is particularly valuable for developers and researchers building knowledge graphs or performing large-scale data analysis. It prioritizes evidence preservation, meaning every extracted fact remains linked to its source text. This focus on transparency and normalization makes it an essential component for any workflow requiring reliable, verifiable structured facts from messy, real-world documentation.
Information Extraction Use Cases
- Converting long-form research papers or reports into queryable knowledge graph triples.
- Automating the population of databases from meeting transcripts and interview notes.
- Extracting structured event timelines and participant roles from news articles or historical documents.
- Normalizing entity mentions across disparate text sources to eliminate data duplication.
- Generating TSV or JSONL datasets for training machine learning models or downstream data analytics.
How Information Extraction Works
- Scope Definition: The system identifies the extraction boundaries and the required level of data granularity.
- Segmentation: Input text is broken down into manageable processing units like sentences and paragraphs.
- Entity Recognition: Primary entities are identified along with specific evidence spans from the text.
- Fact Extraction: The pipeline identifies relations, attributes, and complex events involving the discovered entities.
- Normalization: Aliases and predicates are resolved to a canonical form to ensure consistency.
- Triplication: The data is mapped into triples and formatted for the final output.
- Validation: Ambiguities are flagged for human or secondary AI review to ensure high confidence scores.
Information Extraction Setup
To begin using this tool in your Openclaw Skills environment, ensure you have Python 3 installed. You can process text via direct strings or file inputs.
# Basic extraction from a text string
python3 skills/information-extraction/scripts/extract.py --text "OpenAI published GPT-4." --output out.json
# Pipeline normalization to resolve aliases
python3 skills/information-extraction/scripts/normalize.py --input out.json --output normalized.json
# Exporting to a specific format (JSON, JSONL, or TSV)
python3 skills/information-extraction/scripts/export_triples.py --input normalized.json --format tsv --output triples.tsv
Information Extraction Data Schema & Taxonomy
The skill utilizes a rigorous internal schema to maintain data integrity throughout the extraction lifecycle.
| Component | Description | Key Fields |
|---|---|---|
| Entity | Unique objects or concepts | id, mention, canonical_name, type, confidence |
| Relation | Links between two entities | subject, predicate, object, evidence |
| Attribute | Specific properties of an entity | entity_id, attribute, value, evidence |
| Event | Complex occurrences with participants | id, type, trigger, participants, time |
| Triple | Flattened fact for export | subject, predicate, object |
Information Extraction Advanced Features
- Evidence Mapping: Automatically preserves the exact text span used as evidence for every extracted entity and relation.
- Taxonomy Alignment: Supports controlled predicates through integration with custom relation taxonomies.
- Ambiguity Logging: Instead of guessing, the skill records unresolved ambiguities in a dedicated section for manual verification.
- Multi-Format Export: Native support for JSON, JSONL, and TSV formats to fit various database and graph ingestion workflows.
- Modular Pipeline: Separate scripts for extraction, normalization, and export allow for flexible automation hooks.
SKILL.md
Loading
Related Openclaw Skills
Telegram Send File
A specialized skill for delivering local files, remote URLs, and Telegram file IDs to specific chats and topics via the Bot API.
Code Dev Pipeline
A professional 8-agent collaborative coding pipeline designed for complex software development, frontend design, and automated quality assurance.
HTML to Markdown
A versatile utility to transform HTML code, local files, or live web pages into clean Markdown formatting for AI agents and developers.
War Room Skill
A high-stakes coordination protocol for managing multi-agent teams during production outages and critical system failures.
vLLM High-Performance Inference Engine
An expert assistant for deploying and optimizing large language models using the high-throughput vLLM inference engine.
Market Intelligence Claw
A real-time strategic intelligence layer that transforms live market data into actionable business insights for ecommerce and digital brands.
Featured*
