Abstract:
In this era of internet technology, in Bangladesh, online marketing or e-commerce businesses were already thriving. Due to the COVID-19 pandemic, as people are in lockdown, online shopping became the main platform for shopping as it is the safest way. It accelerated the businesses to come online. More online product service providers makes it better for people but also raises the question of product quality and services. So it is easy for new customers to get scammed while shopping online. Our goal is to make a system that will analyze the customer’s feedback from online shopping and provide a ratio of the positive and negative feedback written in Bangla from the previous customers using Natural Language Processing (NLP). We have collected over 1000 feedback and comments on the product to conduct the research. We used sentiment analysis along with some classification algorithms like KNN, Decision Tree, Support Vector Machine (SVM), Random Forest and Logistic Regression. With the highest accuracy of 88.81%, SVM outperformed all the other algorithms.
