PERSONALIZATION OF WEB SEARCH USING AUTOMATICALLY GENERATED USER PROFILE BASED ON ONTOLOGY

Abstract

The World Wide Web is a collection of billions of documents and web pages. These documents and web pages are increasing day by day. It is very difficult to find relevant information for a particular user on web. Therefore, there is a need of personalized web search system which can retrieve most relevant search results based on user preferences. Various approaches for personalization of web search have been proposed, but it is unclear whether personalization is working efficiently and effectively for different users on different queries with different search contexts. We propose a personalized search approach that works by creating click based user profile. We resolve the ambiguity of the search query before submitting to the search engine. The ambiguity of search query is resolved with the help of user profile, ontological structure and WordNet database. The user profile is also used to personalize web search results obtained from search engine. The main advantage of this approach is that it is not only independent of search engine but also more efficient than common web search. By efficiency, we mean the number of relevant documents retrieved. We find that our approach retrieves more number of relevant documents as compared to previous approach. We also compare the precision and recall for our approach against that of previous approach. The implementation of the proposed approach is done using client-server architecture with help of Java programming language. Users interact with our application with the help of a web browser. Web browser sends user request to Web Server and presents a response from the server to the user