论文信息 - Assessing Systematic Topic Difficulty Based on Query and Collection Features.

Assessing Systematic Topic Difficulty Based on Query and Collection Features.

The National Institute of Standards and Technology (NIST) initiated the Text Retrieval Conference (TREC) in 1990 to evaluate the performance of different retrieval systems and promote the development of text retrieval technology. Intensive experimental studies have been conducted to compare the performance of different systems ever since. A number of influential retrieval systems were able to be forged and enhanced in the conference. The effectiveness of different retrieval systems was compared based on several systematic evaluation measures (e.g. precision, recall or average precision) that were derived from classic Cranfield tests (Cleverdon, Mills & Keen, 1966).

Kun Lu | Kun Lu

[1] Falk Scholer,et al. Effective Pre-retrieval Query Performance Prediction Using Similarity and Variability Evidence , 2008, ECIR.

[2] W. Bruce Croft,et al. Predicting query performance , 2002, SIGIR '02.

[3] W. Bruce Croft,et al. Query performance prediction in web search environments , 2007, SIGIR.

[4] Michael I. Jordan,et al. Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[5] Cyril W. Cleverdon,et al. Factors determining the performance of indexing systems , 1966 .

[6] Oren Kurland,et al. Predicting Query Performance by Query-Drift Estimation , 2009, ICTIR.

[7] Chris Buckley,et al. New Retrieval Approaches Using SMART: TREC 4 , 1995, TREC.

[8] Thomas Hofmann,et al. Probabilistic latent semantic indexing , 1999, SIGIR '99.

[9] Iadh Ounis,et al. Inferring Query Performance Using Pre-retrieval Predictors , 2004, SPIRE.

[10] T. Landauer,et al. Indexing by Latent Semantic Analysis , 1990 .

[11] Donna K. Harman,et al. Overview of the Reliable Information Access Workshop , 2009, Information Retrieval.