Inferring User Search Goals Using Feedback Session

The aim of topic is to discover the number of different user search goals for a query and representing each goal with some s. We first infer user search goals for a query by clustering feedback sessions. For that, we use a concept of pseudo document, which is the revised version of feedback session. At the end, we cluster these pseudo-documents to infer user search goals and represent them with some s. Since the evaluation of clustering is also an important problem, we used evaluation criterion classified average precision (CAP) to evaluate the performance of the restructured web search results. The clustering is done by bisecting k means where in the existing system it is done by k means clustering. The new algorithm increases the efficiency of result. After the segmented result formation, the result in the every segment is reorganized as per number of clicks of URLs. The link which is clicked more number of times will appear at first location in the segment. This reduces the time requirement for searching.