Social Network Service (SNS) is a popular and powerful Web service to connect and/or find friends, and the tendency of a user's interest often affects his/her friends who have similar interests. We can 'Collective Intelligence and Social Big Data' through the SNS. If we can track users' preferences in certain boundaries in terms of Web searching, we can we can improve search efficiency and reliability in view of users. In this paper, we analyze correlation - between Social Relation Value and Search Pattern of web users to improve search efficiency and. Social Relation Value indicates degree of relation on latent characteristic among web users who exist in Social Networks. We calculate Social Relation Value in granting association between 'M' topics and 'L' Attributes that are applicable to 'N' Factors(Similarity, Adjacency, Etc.). And then, we analyze the Correlation - between Social Relation Value and Search Pattern by comparing the search result of each web users correspond to input 'Query' based on Social Relation Value calculated according to topics. Consequently, we find that Search Pattern is similar to that of web users who have high Social Relation Value. Namely, web users who have similar characteristic and interact with each other with activity in (Online) Social Networks have high Correlation. Thus, we are able to improve search efficiency and reliability when the correlation is applied to search.