Understanding a big document by reading it, or finding intended information from a diverse and vast collection of information from online sources is the most time consuming and most tedious task in today era of information explosion. The solution to this problem is to find a summary of the document that explains the important aspects of the document in brief. Trivial approaches to obtain a summary of the document are based on Clustering, Text Ranking and Lexical Chains generate summary but fail to address duplicates problem. The proposed method makes use of latent factors of weighted algorithms and finds topics. And the topics frame the base for building summary.