Sentiment Analysis on Twitter Data using R

Abstract

Nowadays social networking sites are at the blast from where huge amount of information is produced or retrieved. 90% people of the world are sharing their perspectives every day on micro blogging sites, since it contains short and simple expressions. The various devices, mobiles, laptops, tabs and other IoT data gadgets generate huge volume of data and Microservices based web applications running on these have made it simpler for us to get any kind of data at any time and from any place. Social media is also used for expressing our opinions for the products and services. The feedbacks and ratings of millions of the social site users can be collated to extract their attitudes and sentiment towards any products or services and use that information for future market and business improvement or domain analysis. Mining user’s opinion from social media is a difficult task; it can be refined into numerous ways. In this paper, an open source approach is presented which we have collected tweets from Twitter API and then pre-processed, analyzed and visualized these tweets using R. To analyze sentiments of tweets we are utilizing a statistical tool, R programming. This sentiment analysis is based on text data retrieval from streamed web and then classifying people perspectives in eight distinct classifications of feeling (disgust, fear, anger, anticipation, sadness, trust, surprise) and two unique sentiments (positive and negative).