![]() ![]() Ghose, A., Ipeirotis, P.G.: Estimating the helpfulness and economic impact of product reviews: mining text and reviewer characteristics. Based on the overall performance, LDA is considered the best topic modelling method for this study. The optimum number of topics for both LDA and LSI is 3 with a coherence score of 0.50451438 and 0.4573542, respectively. As topic modelling techniques, LDA and LSA are utilized to detect, compare and evaluate the major topics within the dataset. In pursuit of this goal, the study focuses on 844 Goodreads reviews of the Iranian diasporic writer Firoozeh Dumas’s Funny in Farsi to analyze the variety of topics that are covered in readers’ responses. This study aims to perform text mining techniques-opinion mining using topic modeling-to examine the range of topics that are explored within reader reviews in social media. The rapid growth of the Internet and the social networking services (SNS) has allowed readers from across the globe to share their thoughts and feelings about literary works through social media platforms like Twitter, Goodreads, and Facebook, to name only a few. ![]()
0 Comments
Leave a Reply. |