Related search recommendation with user feedback session

Abstract

Keyword-based search is an extensively used method to discover knowledge on the Web. Generally, Web users are unable to arrange and define input queries relevant to their search because of adequate knowledge about the domain. Therefore, the input queries are normally ambiguous and short. Query suggestion is a method to recommend queries related to the user input query that helps them to locate their required information more precisely. It helps the search engine to provide relevant answers and meet users needs. Usually, users query keywords are ambiguous, therefore it is not good to use users query keyword in suggestion. In this chapter, Related Search Recommendation (RSR) framework is presented that determines keywords presented in un-clicked and clicked documents in the feedback session. Feedback sessions are used to retrieve users need in terms of Pseudo documents. Semantic similarity is computed between the terms in the Pseudo document. The semantic terms are used for suggestions. The presented method provides semantically related search queries for the user input query. Results show that the RSR method outperforms Rochios model and Snippet Click Model.