Finding dimensions for queries

Abstract

We address the problem of finding multiple groups of words or phrases that explain the underlying query facets, which we refer to as query dimensions. We assume that the important aspects of a query are usually presented and repeated in the query's top retrieved documents in the style of lists, and query dimensions can be mined out by aggregating these significant lists. Experimental results show that a large number of lists do exist in the top results, and query dimensions generated by grouping these lists are useful for users to learn interesting knowledge about the queries.

Faceted Search using Bootstrap List Extraction Method

“…In addition, the performance of related faceted search methods and techniques in all phases is evaluated and described. Dou et al [4] developed QDMiner system to automatically extract facet hierarchies for keywords by aggregating frequent lists from free text, HTML tags, and reiterated regions within top search results. Anju G R and Karthik M [5] proposed a system implementing graphical model which is a supervised method.…”

Extracting query facets from search results

“…A valid query facet that describes Mars landing news could be {cnn, abc, fox }, which includes different news channels. In a recent work [7], Dou et al developed a system to extract query facets from web search results and showed the potential of doing so. However, the unsupervised method they proposed is far from optimal, and it does not improve by having human labels available.…”

“…Faceted search is similar to query facet extraction in that both of them use sets of coordinate terms to represent different facets of a query. However, most existing works for faceted search are build on as specific domain or predefined categories [7], while query facet extraction does not restrict queries in a specific domain, like products, people, etc.…”

“…Similar to Dou et al [7], we use pattern-based semantic class extraction approach [25] to extract lists of coordinate terms from search results as candidates for query facets. In pattern-based semantic class extraction, patterns are applied on the corpus to discover specific relationships between terms.…”

“…Besides lexical patterns, HTML patterns are often used on HTML documents to extract coordinate terms from some HTML structures, like <UL>, <SELECT> and <TABLE>. We use both of the two types of patterns, summarized in [7], lists from each column and each row are extracted.…”