Measuring the semantic similarity between words is an important component in various tasks on the web such as relation extraction, community mining, document clustering, and automatic metadata extraction. The optimal combination of page count based word co-occurrence measures and the lexical pattern clusters are learned using support vector machine. The personalized web search represents the user's search intention and the related concepts based on the user interest. The personalized search is used to display the user's search intention as positive links and the general profiles are displayed as a negative links. The RSVM is used for displaying the top ranking results at first based on the f-score method.