What it does This repository is a structured index of conference and journal papers on automated fact-checking, maintained by the authors of two survey papers (TACL 2022 and EMNLP 2023). It organizes research into a three-stage pipeline: claim detection, evidence retrieval, and claim verification, with a newer multimodal extension. Each section links to datasets, models, shared tasks, and tutorials.

The interesting bit The taxonomy itself is the product. Rather than dumping PDFs, the authors impose a framework that unifies scattered terminology across NLP, computer vision, and social media research — then keep it current with new conferences. The recent additions for LLM factuality and generated-text detection show the taxonomy evolving with the field.

Key highlights

• Covers text, multimodal, and emerging LLM-related fact-checking research
• Explicit task definitions with visual pipeline diagrams
• Datasets tagged by year and venue with direct links to papers and code
• Sections on related but distinct tasks: stance detection, fake news, adversarial attacks
• Active maintenance: updated through NeurIPS and EMNLP 2024

Caveats

• No search or filtering beyond manual browsing; finding specific papers means scrolling
• README is thorough but long; the signal-to-noise ratio favors comprehensive coverage over quick lookup

Verdict Researchers entering fact-checking or survey authors needing a structured starting point should bookmark this. Practitioners looking for off-the-shelf tools or code will find mostly pointers, not implementations.