Researchers have introduced GreekBarRetrieval, a new benchmark designed to improve statutory retrieval for legal question answering in Greek. This benchmark, derived from the existing GreekBarBench, includes 283 bar exam questions and over 6,000 candidate statutory articles. Experiments showed that while dense retrieval methods initially outperformed sparse methods like BM25, an LLM-based query reformulation technique, particularly a ten-round ReAct-like loop, significantly improved BM25's performance and achieved the best overall scores for nDCG and MAP. AI
IMPACT Enhances LLM capabilities in specialized legal domains, potentially improving access to legal information.
RANK_REASON The cluster describes a new academic benchmark and associated research paper published on arXiv. [lever_c_demoted from research: ic=1 ai=1.0]
Read on arXiv cs.IR (Information Retrieval) →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
