Cluster analysis or clustering is the process of assigning a set of objects into groups such that the objects in the same cluster are more similar to each other and objects in other clusters are dissimilar. In the existing system, Locality preserving indexing method is used. Euclidean distance is used as the method to find the distance between the documents. This method also takes use of weighted function to find the distance between documents. Major difficulty is found in analyzing the weighted function. Since the documents can take many forms, the structure of the document space depends on the similarities between the documents. A new approach based on Correlation as a similarity measure is best proposed for analyzing the structure of documents arranged in the high-dimensional document space.