A Link-click-concept Based Ranking Algorithm for Ranking Search Results

Personalized search is an essential research area that has main goal to determine the uncertainty of query terms. In order to enhance the relevance of search results, personalized search engines form user profiles which capture the users' personal preferences and by using those preferences find out the actual goal of the input query. By using User profile we can rank the documents in a search engine according to the query which is submitted by user. A better user profiling strategy is an important and primary component in search engine personalization. In this work, we propose a scheme that supports mining a user's conceptual preferences from users' click through data resulted from web search. This discovered preference is helpful to adapt a search engine's ranking function. In the system, an absolute set of conceptual preferences is derived for a user such as the concepts extracted from the search results and the click through data. After that, a Concept-based User Profile (CUP) in other words a concept ontology tree is generated. Our system proposes a novel approach such as Link-Click-Concept based Ranking Algorithm. The proposed system considers the concept for the user profile construction and has high efficiency than the existing system.