Sentiment Analysis and Product Review Classification in E-commerce Platform

Abstract

Online shopping is becoming one of the most de-manding everyday needs, nowadays. These days people are feeling comfortable shopping online. The number of its customers is increasing day by day as well as raising some problems. The major problem is that the customers can not choose the quality-full product by reading every review of an online product. Besides, the product reviews are helpful to improve the services of an e-commerce site but required huge manpower and time. We have focused on Bangla text and aimed to solve these problems by the application of Deep Neural Network (DNN) and Natural Language Processing (NLP). In this study, we have proposed two deep learning NLP models: one is for sentiment analysis and the other one is for Product Review Classification intended to improve both the quality and services. Significantly, our proposed models result in high accuracy: 0.84 and 0.69 for both Sentiment Analysis and Product Review Classification, respectively. Undoubtedly, these models can help the customers to choose the right product and the service provider to improve their services.