Bangla Product Review Sentiment Analysis By ML & DL Approach

Abstract

In this age of internet technology, e-commerce and online marketing companies in Bangladesh were already doing well. As a result of the widespread quarantines caused by the COVID-19 pandemic, online shopping has surpassed all other methods as the preferred method of making purchases. As a result, Businesses were able to get online much more quickly. While an increase in online product service providers does many good things, it also brings up questions about the reliability and quality of these offerings. Consequently, new customers are easy prey for internet scammers. Our goal is to develop an AI system that can automatically analyze customer reviews of e-commerce products and provide a set of positive and negative comments made by previous customers in the Bangla language using Natural Language Processing (NLP) and Artificial Intelligence algorithms. From different e-commerce websites, we collected 1016 Bangla product review comments for analysis. After completing preprocessing, we used two different domains of AI called Machine Learning(ML) and Deep Learning(DL) algorithms. There are five ML algorithms like KNN, Random Forest, Logistic Regression, SVM, and Gaussian Naïve Bayes used and for deep learning we used CNN and LSTM.