Recommendation Model Using User Reviews and Deep Learning Approach
Abstract— Due to the growing amount of information on the web, the information overload problem arises for users to discover relevant products easily and quickly. Collaborative Filtering (CF) is a recommender system which uses the user-item rating matrix to make suggestions to users by finding many users’ similar preferences. User reviews have information about products and users’ opinions than ratings. Using Opinion-based ratings in the rating prediction process of recommender system can reflect the detail preference of users effectively and accurately. In this paper, sentiment analysis is used to find the opinions of the user and matrix factorization that applied neural network is used to predict the ratings for the recommendation. The experiment is performed on the dataset of the hotel reviews and rating prediction accuracy is evaluated by measuring the prediction error called the Root Mean Square Error (RMSE).
Index Terms— Collaborative Filtering, Opinion-based ratings, sentiment analysis, neural network
Nyein Ei Ei Kyaw, Thinn Thinn Wai
University of Information Technology, Yangon, MYANMAR
Cite: Nyein Ei Ei Kyaw, Thinn Thinn Wai, "Recommendation Model Using User Reviews and Deep Learning Approach, " Proceedings of 2021 the 11th International Workshop on Computer Science and Engineering (WCSE 2021), pp. 49-53, February 25-27, 2021.