A new arXiv preprint argues that the two most common query-document scoring methods, TF-IDF and BM25, both correspond exactly to Kullback-Leibler divergences. This gives the widely used retrieval heuristics a probabilistic justification and places them inside a single statistical framework. The author frames the work as filling a long-standing gap in how these ranking functions are derived.
A new arXiv paper introduces SatIR, a retrieval method aimed at settings where candidates must meet a specific profile's constraints rather than just be topically similar. Clinical trial matching is used as the motivating high-stakes case, where missing an eligible patient or trial carries real cost. The approach targets scalable, high-recall constraint satisfaction across many competing profiles.
A new arXiv paper introduces CompCQR, a method that rewrites ambiguous, context-dependent user utterances into clearer retrieval queries for multi-turn conversational search. The approach composes queries without requiring task-specific training, aiming to bridge the gap between conversational phrasing and standard retrieval systems. The work targets information-seeking dialogue, where follow-up questions often lack the context needed for direct use as search queries.