Comparison between Multinomial Naive Bayes and Multi-Layer Perceptron for Product Review In Real Time

To find out and understand users’ requirements, complaints, and preferences, user feedback can be used as a perfect tool. Taking into account the amount of user feedback received on a product every day, it is difficult for the developers to identify issues that users have for the product or services from the user-given feedback, as it takes a lot of time and thus becomes very inefficient. To overcome this problem, an automation system is used to fetch the real-time product reviews and then a comparison between Multinomial Naïve Bayes and Multi-Layer Perceptron was built to determine which algorithm performs better with a dataset of a real time product review. Document/Text Classification and language representation in vector form is useful in the field of machine learning. In this paper, we aim to use the Text Classification and vector representations method to decode the reviews gathered in real-time into supervised form and then apply the Multinomial Naïve Bayes and Multi-Layer Perceptron classification algorithm to evaluate the Sentiment Analysis on the website products.