Personalized Web Search by Mapping User Queries to Categories by Positive Preferences

User profiling is a fundamental component of any personalization applications. The users interests are performed based on user profiling method (i.e., positive preferences), but not the objects that users dislike (i.e., negative preferences). The paper, mainly focus on the users both positive and negative preferences that is depend upon many search engine personalization and develop several concept-based user profiling methods. The previously proposed evaluation method is personalized query clustering method. The proposed Experimental results show that a profile which capture and utilize both of the users's positive and negative a preference which performs best. An important result from the experiments is that profiles with negative preferences can increase the time of query cluster and at the same time separation between similar and dissimilar queries. The separation provides a clear threshold for an agglomerative clustering algorithm to terminate and improve the overall quality of the resulting query clusters. The proposed system used to producing the relevant information to the top of the search page. The letter pair algorithm is used for proposed system. This algorithm is mainly used for matching the keyword with profile. The  keyword is also called tags. This method extracts the result from web-snippets and clustering them. Then the users interests are displayed the front of the search page.