Sentiment Analysis of Online Reviews

Sentiment Analysis of Online Reviews

Author Name : Deepika.M, Mr. Y. Ramu

ABSTRACT

Ideas and opinions from other people has a great chance of impact in changing our opinions. With the advent expansion of digitalization modern era is entirely loaded with data and opinions where people express their opinions, emotions and attitudes in the form of comments, user reviews, posts, feedback, tweets about certain brands, events, services, products or company through written language in online platforms. Thus at the end, people have become increasingly interested in mining the massive amounts of information to find insights. Opinion mining and sentiment analysis are two computational methods of analyzing sentiments, subjective opinions, and attitudes in a textual messages. Sentiment analysis is a process of determining the sentiment from a user feedback or textual message. Where as sentiment is the one that tells whether the textual data is positive or negative. It decides whether the textual data is a positive opinion or negative opinion. Sentiment analysis of online reviews is about prediction of sentiment form the online user reviews. The most important thing to focus on is determining if the opinion expressed by the author is negative or positive. Almost every online websites facilitate the users to give comments, feedback or review for the services offered by them and a huge volume of information is displayed. Taking out the relevant data found in users reviews, provides us a variety of applications in various domains. In order to analyze online sentiments, we propose a machine-learning models for predicting overall sentiments across online reviews. As a part, we have employed algorithm viz Multinomial Naive Bayes, Random forest and predicted the sentiment of review in our proposed work.

Keywords: Sentiment Analysis, Approaches, Sentiment Prediction, Multinomial Naïve Bayes, Random Forest,