TY - JOUR
AU - Burges, Christopher J.C.
AB -  A Machine Learning Approach for Improved BM25 Retrieval Krysta M. Svore Microsoft Research One Microsoft Way Redmond, WA 98052 Christopher J. C. Burges Microsoft Research One Microsoft Way Redmond, WA 98052 ksvore@microsoft.com cburges@microsoft.com ABSTRACT Despite the widespread use of BM25, there have been few studies examining its e €ectiveness on a document description over single and multiple eld combinations. We determine the e €ectiveness of BM25 on various document elds. We nd that BM25 models relevance on popularity elds such as anchor text and query click information no better than a linear function of the eld attributes. We also nd query click information to be the single most important eld for retrieval. In response, we develop a machine learning approach to BM25-style retrieval that learns, using LambdaRank, from the input attributes of BM25. Our model signi cantly improves retrieval e €ectiveness over BM25 and BM25F. Our data-driven approach is fast, e €ective, avoids the problem of parameter tuning, and can directly optimize for several common information retrieval measures. We demonstrate the advantages of our model on a very large real-world Web data collection. Categories and Subject Descriptors: H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval I.2.6 [Arti 
TI - A machine learning approach for improved BM25 retrieval
DO - 10.1145/1645953.1646237
DA - 2009-11-02
UR - https://www.deepdyve.com/lp/association-for-computing-machinery/a-machine-learning-approach-for-improved-bm25-retrieval-SL0dEunCgA
DP - DeepDyve
ER -