Sentiment Analysis on E-Commerce Website

1Chaitanya Saha, Vishal Singh, S.S. Sridhar

125 Views
20 Downloads
Abstract:

The growth of sentiment analysis is a major achievement in Computer Science. Sentiment analysis is the technique to gather the text data and analyze the data for getting polarity differences in the sentences. It helps to analyze the opinions, emotions etc. that are exchanged by and between humans. The data can be divided into different segments depending upon the meaning of the words. The data is growing rapidly and undoubtedly it is a rich source of information. Thus Sentiment analysis can be useful for analysis in various fiends such as analysis of the comments or views on social media, politics, disaster management etc. In this paper, we are doing Sentiment Analysis on the reviews of e-commerce website to understand the mood and opinion of the people by analyzing the reviews given by them on the website by using various NLP techniques and NLTK along with other processes. We are trying to build a system that can segregate the sentences into positive sentences which have no abhor discourse and negative sentences which can be either hateful or offensive. We try to analyze the scarcasm and emotions too. Sentiment of a review tends to analyses the brand value that lead to further improvise of product. In conclusion we get an approximate accuracy of 87% and by this we can infer that the system will help to improve the overall decision making process for the brands and the customers.

Keywords:

sentiment, NLP, NLTK, abhor discourse

Paper Details
Month4
Year2020
Volume24
IssueIssue 8
Pages1071-1079