Researchers have introduced Sophia-bench, a new benchmark designed to evaluate patent retrieval systems across diverse real-world search scenarios. This benchmark includes 10,000 queries and 75,000 documents, covering various technology sections and jurisdictions, and assesses retrieval performance using 12 different query types. Alongside the benchmark, they developed QaECTER, a 344M-parameter embedding model that achieves state-of-the-art results in patent retrieval, outperforming larger models on Sophia-bench and other independent evaluations. AI
Summary written by None from 1 source. How we write summaries →
IMPACT Establishes a new benchmark and model for patent retrieval, potentially improving innovation and IP strategy decision-making.
RANK_REASON The cluster describes a new academic paper introducing a benchmark and a model for patent retrieval.