Exploiting User Queries for Search Result Clustering

摘要

Search Result Clustering (SRC) groups the results of a user query in such a way that each cluster represents a set of related results. To be useful to the user, the different cluster should contain the results corresponding to different possible meanings of the user query and the cluster labels should reflect these meanings. However, existing SRC algorithms often ignore the user query and group the results based just on the similarity of search results. This can lead to two problems: low quality cluster, where the results within a single cluster are related to different meanings of the query; and poor cluster labels, where the label of the cluster does not reflect the query meaning associated with the results in the cluster. This paper presents a new SRC algorithm called QSC that exploits the user query and uses both syntactic and semantic features of the search results to construct clusters and labels. Experiments show that the query senses are good candidates for the cluster labels and the algorithm can lead to high quality cluster and more semantically meaningful labels than other state-of-the-art algorithms.