High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Recommendation System Depends on User Comments Based Product Rating by Sentiment Analysis

Author(s):

B. Ghowsalya , PSR RENGASAMY COLLEGE OF ENGINEERING FOR WOMEN, SIVAKASI; R. Kalaivani, PSR RENGASAMY COLLEGE OF ENGINEERING FOR WOMEN, SIVAKASI; Dr. C. Balasubramaniyan, PSR RENGASAMY COLLEGE OF ENGINEERING FOR WOMEN, SIVAKASI

Keywords:

Information Overloading; Rating System; Score Forecast; Topic Modeling

Abstract

Nowadays we have many websites that allows user to share their viewpoints about a particular product. This makes a great opportunity to know what originally a user feels about a product. But there they are facing information overloading problem. From the available tremendous amount of reviews extracting valuable reviews is very difficult. The previous rating system consider some factors like customers purchase record, manufactured goods category, users geographical location not considering user comments. In this work, we are proposing a score forecast based on sentiment analysis that improve accuracy in rating systems. Firstly we propose a topic modeling approach calculate LDA score to find out users preferred topics. Secondly, we are producing a score forecast approach to calculate score, based on the user commands. This process will improve the rating prediction accuracy in recommender systems.

Other Details

Paper ID: IJSRDV6I20196
Published in: Volume : 6, Issue : 2
Publication Date: 01/05/2018
Page(s): 3714-3717

Article Preview

Download Article