Improved Similarity Based Matrix Completion Algorithm Using Latent Semantic Indexing

Abstract—LSI usually is conducted by using the singular value decomposition (SVD). The main difficulty in using this technique is its retrieval performance depends strongly on the choosing of an appropriate decomposition rank. In this paper, by observing the fact that the SVD makes the related documents more connected, we devise a improved matrix completion algorithm. The proposed algorithm returns results that are meaningful to the search criteria. Latent Semantic Indexing (LSI) helps in information filtering where certain type of words are removed from retrieved documents and indexing is performed to bring the essential results at the top and so on according to the defined algorithm. I