Deep Fact Validation
FEAT, short for Factcheck Explorer Analysis Tool, is designed to facilitate the exploration, analysis, and visualization of fact-checking data.
A self-hosted data validation platform, for labor intensive fact checking.
A Benchmark Dataset for Multimodal Scientific Fact Checking
A lightweight, agent-style framework for fact-checking atomic claims using iterative retrieval and verification. Reduces LLM and search cost while maintaining strong factuality performance.
Misinformation papers @ CSCW 2018 & ICWSM 2020.
Android App to help you check claims from the internet, WIP
API for the browser extension. The API computes scores and stores result in a DB
Application that automatically transcribes and verifies factual claims from videos, Instagram posts, and free-text input using OpenAI’s most advanced models. Designed for seamless, multilingual fact-checking with web-integrated validation and intuitive UX.
Displays a confidence score about an article that has been submitted to the extension.
Low Resource Context Relation Sampler for contexts with relations for fact-checking and fine-tuning your LLM models, powered by AREkit
The Reality Button is a Chrome extension that provides on-demand, AI-powered verification of selected text anywhere on the web. Users can highlight a claim, right-click, and receive a concise assessment of its factual accuracy, including a verdict, confidence score, explanation, and supporting sources.
Project Mirador codebase and Stelthar API
This project aims to implement text classification architecture using pre-trained language model BERT.
Add factchecks to your laravel application
Website in which you can scroll through factchecks that have been scraped from newssites. Created with React. Uses an API that I've made with Python.
To associate your repository with the factchecking topic, visit your repo's landing page and select "manage topics."