Mining Facets for the Searched Queries

Abstract

We deal with the problem of discoveringquery facets which are several groups of words or phrases that make clear and review the content enclosed by a query. We believe that the significant aspects of a query are usually presented and recurred in the query’s peak retrieved documents in the style of lists, and query facets can be mined out by aggregating these important lists. We propose an organized answer, which we refer to as QDMiner, to automatically supply query facets by extracting and grouping recurrent lists from free text, HTML tags, and duplicate regions within top search results. Experimental outcome show that a big number of lists are present and valuable query facets can be mined by QDMiner. We further analyze the problem of list duplication, and find superior query facets can be mined by modeling fine-grained similarities between lists and punishing the duplicated lists.