SearchPlus: a new search results clustering approach

Search engines are the fastest way to find and retrieve information on the web. Their usefulness is evident from the fact that three of the most visited web sites on internet are search engines. With advanced algorithms and the ability to go over millions of records in mere milliseconds, the most time consuming part of searching for the required information is not the search itself but the time it takes to sort through the searched results. Specially when the search query is vague, too broad or simply ill defined. This leads the user to a situation where the search results becomes unmanageably large and confusing. In this paper we present a novel search result clustering technique that relies on the information returned by the search query to extract keywords and to cluster the search results. Unlike most of the search result clustering techniques our proposed solution does not rely on any external knowledge bases, predefined categories or any peripheral knowledge. Thus our technique could be applied on any search domain. We developed a prototype to conduct experiments. Our preliminary results show the applicability and effectiveness of our approach.