Dynamic Query Clustering in Personalized Search to Improve Retrieval Relevance

Today internet users are using web search engines for complex generic query processing to achieve the day to day activities like trip management, budget planning, shopping plan and text similarity etc. To avoid the complex generic query management many search engines are using query modulators, to break the main query in smaller sub queries to reduce the complexity and to extract the relevant data as results. Some search engines also maintain the user level search history customization to help the user by suggesting them. To achieve more efficiency in personalized search, in this paper we are introducing dynamic clustering in personalized search to assist the user search and to improve the precision of search relevance. This clustering is also useful to find result ranking, relevance ratio and result collaboration. Experimental results are showing that our approach is having the high precision and recall in terms of search relevance and scalable in terms of response time than other approaches.

Keywords: web mining, personalized search, dynamic query clustering, search history analysis, query result processing.