E-Commerce Product Review Classification based on Supervised Machine Learning Techniques

To become a successful organization it is the most important aspect to consider is the customer satisfaction. A business can grow only when the customers enjoy their product/service and how much they recommend it. For the same, companies collect reviews and ratings from its customers to understand their perspective better. Amazon is also such a giant, which values its customers’ opinions. It is very difficult or nearly impossible to have a detailed manual scan of each and every review for each and every product available on the website. This is where NLP (Natural language processing) comes into picture. It processes human language from text to arrays of numbers in order to find sense in the text so that the machine learning algorithms can process them further and classify the texts. This project focuses on processing the amazon reviews of various categories and products and then classify them into 5 ratings, from 1 to 5, worst to best respectively. The supervised machine learning algorithms used to process these reviews are Support Vector machines (SVM), Naïve Bayes classifier and Decision Tree classifier. The analysis of comparison of these algorithms has also been discussed using accuracy score bar charts.