Enhanced clustering models with wiki-based k-nearest neighbors-based representation for web search result clustering

The increases in the information resources on the World Wide Web in search of the necessary information, as users navigate the Web with multiple sites. When user surfing the web which is a huge and complicated often miss their required searching pages. Web personalization is based on the Web usage logs of user's makes advantage of the knowledge required for the analysis of the content and structure of web sites promising to solve this problem by supporting one of the procedures. The search engine can affect the effectiveness of existing approaches, depending on the user profile, which is building more and more on the web pages or documents. In this paper, we propose an efficient and novel web search based on the individual classification and clustering method. The proposed approach classified the cluster data using frequent pattern mining and multilevel association rules for recurring relationship and cluster the web usage using Hierarchical methods with the navigating site and user interest for personalization. This approach process in advance to support the real time personalization and minimizes the cost reduction of preparation personalization resource in real time. The proposed approach is an effective personalization to the user's interest; in experimental research it has shown high precision measures.

K. Vijayalakshmi

Papers

Web Usage Classification and Clustering Approach for Web Search Personalization