OpenFactVerification  by Libr-AI

Open-source tool for automated fact verification

Created 1 year ago
1,108 stars

Top 34.5% on SourcePulse

GitHubView on GitHub
Project Summary

Loki is an open-source tool designed to automate fact verification by dissecting text into claims, assessing their verifiability, generating evidence queries, and performing the verification. It targets journalists, researchers, and anyone needing to assess information factuality, offering a comprehensive pipeline for automated fact-checking.

How It Works

Loki employs a modular pipeline architecture. It first breaks down input text into individual claims, then evaluates each claim's suitability for verification. Subsequently, it generates search queries to find supporting or refuting evidence, crawls the web for this evidence, and finally synthesizes the findings to verify the original claims. This structured approach allows for systematic and automated assessment of information.

Quick Start & Requirements

  • Installation: Clone the repository and install dependencies using poetry install or pip install -r requirements.txt within a Python 3.9+ environment.
  • Prerequisites: Requires API keys for services like Serper (for evidence retrieval) and OpenAI (for various tasks). These can be configured via environment variables or a YAML file.
  • Usage: Can be used as a library (from factcheck import FactCheck) or as a web app (python webapp.py --api_config demo_data/api_config.yaml). Supports multimodal inputs including text, speech, image, and video.
  • Documentation: User guide available for advanced usage. Online service demo available.

Highlighted Details

  • Supports multimodal fact verification (text, speech, image, video).
  • Integrates a full pipeline from claim extraction to evidence retrieval and verification.
  • Offers both library and web application interfaces.
  • Requires API keys for core functionalities.

Maintenance & Community

The project welcomes contributions and provides contribution guidelines. Community engagement is encouraged via a newsletter and Discord server.

Licensing & Compatibility

The project's license is not explicitly stated in the README. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The README does not specify the project's license, which may impact commercial adoption. Detailed performance benchmarks or limitations of the verification accuracy are not provided.

Health Check
Last Commit

11 months ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
5 stars in the last 30 days

Explore Similar Projects

Starred by Dan Abramov Dan Abramov(Core Contributor to React; Coauthor of Redux, Create React App) and Edward Sun Edward Sun(Research Scientist at Meta Superintelligence Lab).

LeanDojo by lean-dojo

0%
705
Machine learning for theorem proving in Lean
Created 2 years ago
Updated 5 days ago
Feedback? Help us improve.