Automated Feedback System for Customer Reviews using Sentiment Analysis

Abstract

Today, public reviews in terms of digital nature play a vital role in customer buying pattern and customer satisfaction.People may not be able to read all public/customer reviews of a product/movie and get the relevant feedback.The aim of the paper is to automate the process for providing effective feedback to customers by extracting subjective information from the text or feedbacks.This approach is unique in the following way: in case of positive feedbacks the customer is greeted with an appreciation message else a message conveying that services would be improved is sent to customers.In this proposed research, a model has been developed which performs sentiment analysis by using Natural Language Tool Kit (NLTK).The classification was done using Naive Bayes classifier by calculating the probability of the reviews of the customers with the highest value is considered as either positive or negative.As a result, the emotions and sentiment analysis improves psychology of the customer, reduces the stress and improves product purchase rate.