Paper Details

Nowdays we use many online shopping sites for shopping.In E-commerce applications various trust models areused based on reputation. Peoples purchase products from listed sellers based on their feedback ratings. All majoronline shopping sites encourage buyers to provide feedback, often in the form of ratings along with some textualcomments to facilitate potential transactions. Sometime buyers gives star ratings from 1 to 5 stars for a particularproduct but if he/she is not satisfied about any other issue like about delivery, payment options etc. then he/she givesnegative comment in text field. So it is not trustworthy for other customers to buy that product or not. Also at a timereading all the comments is not possible.To overcome all above issues we proposed model named as “Comm-Trust” based on opinions that user gives in free textcomment field. This model includes :1) Computation of dimension trust scores and dimension weights automatically by extracting dimension ratings fromfeedback comments. 2) An algorithm for dimension rating and weights combining natural language processing, opinionmining, topic modelling. Finally we can display percentages of +ve, -ve, and neutral comments in histogram form

Akshay Zaware, Tushar Kamble, Tushar Kardak, Vyankatesh Hardas, Mrs.Deshmukh P. R.(Assistant Professor), “Computing Multi-dimensional Trust By Mining E-Commerce Feedback Comments”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 4, Issue 5, pp. 829-831, May 2017.