Abstract:Search diversification (also called diversity search), is an important approach to tackling the query ambiguity problem in information retrieval. It aims to diversify the search results that are originally ranked according to their probabilities of relevance to a given query, by re-ranking them to cover as many as possible different aspects (or subtopics) of the query. Most existing diversity search models heuristically balance the relevance ranking and the diversity ranking, yet lacking an efficient learning mechanism to reach an optimized parameter setting. To address this problem, we propose a learning-to-diversify approach which can directly optimize the search diversification performance (in term of any effectiveness metric). We first extend the ranking function of a widely used learning-to-rank framework, i.e., LambdaMART, so that the extended ranking function can correlate relevance and diversity indicators. Furthermore, we develop an effective learning algorithm, namely Document Repulsion Model (DRM), to train the ranking function based on a Document Repulsion Theory (DRT). DRT assumes that two result documents covering similar query aspects (i.e., subtopics) should be mutually repulsive, for the purpose of search diversification. Accordingly, the proposed DRM exerts a repulsion force between each pair of similar documents in the learning process, and includes the diversity effectiveness metric to be optimized as part of the loss function. Although there have been existing learning based diversity search methods, they often involve an iterative sequential selection process in the ranking process, which is computationally complex and time consuming for training, while our proposed learning strategy can largely reduce the time cost. Extensive experiments are conducted on the TREC diversity track data (2009, 2010 and 2011). The results demonstrate that our model significantly outperforms a number of baselines in terms of effectiveness and robustness. Further, an efficiency analysis shows that the proposed DRM has a lower computational complexity than the state of the art learning-to-diversify methods.

A Subtopic Taxonomy-Aware Framework for Diversity Evaluation.

Learning to Diversify Search Results Via Subtopic Attention

Supervised Search Result Diversification Via Subtopic Attention

Multi-dimensional Search Result Diversification

Directly Optimize Diversity Evaluation Measures: A New Approach to Search Result Diversification.

Low-cost, Bottom-Up Measures for Evaluating Search Result Diversification

Microsoft Research Asia at the Web Track of TREC 2009.

Learning To Improve Affinity Ranking For Diversity Search

Improve Web Search Diversification with Intent Subtopic Mining

Result Diversification in Search and Recommendation: A Survey

Revisiting The Evaluation Of Diversified Search Evaluation Metrics With User Preferences

Search Result Diversification Based on Query Facets

Leveraging Integrated Information to Extract Query Subtopics for Search Result Diversification

The Search Result Diversification Approach Based on the HITS Algorithm

Learning to diversify web search results with a Document Repulsion Model.

Simple Evaluation Metrics for Diversified Search Results.

A Survey on Search Result Diversification

A Learning Approach to Hierarchical Search Result Diversification.

Diversified Search Evaluation: Lessons from the NTCIR-9 INTENT Task

A supervised term ranking model for diversity enhanced biomedical information retrieval

Evaluating Search Result Diversity Using Intent Hierarchies.