VERIRAG: Healthcare Claim Verification via Statistical Audit in Retrieval-Augmented Generation

作者：Shubham Mohole, Hongjun Choi, Shusen Liu, Christine Klymko, Shashank Kushwaha, Derek Shi, Wesam Sakla, Sainyam Galhotra, Ruben Glatt · 年份：2025 · DOI：10.1145/3765612.3767788 · 被引用次数：2 · 研究领域：Meta-analysis and systematic reviews、Topic Modeling、Biomedical Text Mining and Ontologies

Retrieval-augmented generation (RAG) systems retrieve clinically-relevant evidence but remain methodologically blind, unable to judge study quality (e.g., retractions, underpowered analyses). We introduce VERIRAG, which addresses this gap through a three-part framework: (i) an 11-point Veritable audit for methodological rigor; (ii) a quality- and novelty-weighted Hard-to-Vary (HV) score to aggregate evidence; and (iii) a Dynamic Acceptance Threshold calibrated to claim boldness. Across four corpora simulating the evolution of scientific evidence, from initial flawed findings (TY0) to settled science (TY5), VERIRAG consistently outperforms RAG baselines like COT-RAG[2], Self-RAG[1], FLARE[4], and CIBER[3].