In this paper, the authors present k - means clustering algorithm used to find out the ranking from given user information available on social network web sites like orkut, facebook, twitter and collaborative filtering algorithm for reducing information overload. It is one of the simplest clustering algorithms. It is called k-means because it iteratively improves the authors' partition of the data into k sets. Then find out the average of each dataset (data mining value). This algorithm is used to reduce the work complexity.