Loading...
Please wait, while we are loading the content...
Similar Documents
Query-Performance Prediction: Setting the Expectations
| Content Provider | CiteSeerX |
|---|---|
| Author | Raiber, Fiana Kurland, Oren |
| Abstract | The query-performance prediction task has been described as estimating retrieval effectiveness in the absence of rele-vance judgments. The expectations throughout the years were that improved prediction techniques would translate to improved retrieval approaches. However, this has not yet happened. Herein we provide an in-depth analysis of why this is the case. To this end, we formalize the pre-diction task in the most general probabilistic terms. Using this formalism we draw novel connections between tasks — and methods used to address these tasks — in federated search, fusion-based retrieval, and query-performance pre-diction. Furthermore, using formal arguments we show that the ability to estimate the probability of effective retrieval with no relevance judgments (i.e., to predict performance) implies knowledge of how to perform effective retrieval. We also explain why the expectation that using previously pro-posed query-performance predictors would help to improve retrieval effectiveness was not realized. This is due to a mis-alignment with the actual goal for which these predictors were devised: ranking queries based on the presumed effec-tiveness of using them for retrieval over a corpus with a spe-cific retrieval method. Focusing on this specific prediction task, namely query ranking by presumed effectiveness, we present a novel learning-to-rank-based approach that uses Markov Random Fields. The resultant prediction quality substantially transcends that of state-of-the-art predictors. |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Query-performance Prediction Retrieval Effectiveness Effective Retrieval Markov Random Field Resultant Prediction Quality Query-performance Pre-diction Query Ranking Fusion-based Retrieval Presumed Effec-tiveness Pre-diction Task Query-performance Prediction Task Actual Goal Presumed Effectiveness Novel Connection Novel Learning-to-rank-based Approach Formal Argument Spe-cific Retrieval Method Pro-posed Query-performance Predictor Specific Prediction Task Federated Search Improved Prediction Technique Rele-vance Judgment In-depth Analysis General Probabilistic Term Retrieval Approach State-of-the-art Predictor Relevance Judgment |
| Content Type | Text |
