Evaluating Sentiment Analysis Techniques for Political Tweet Data

Authors

  • Kirti M. Jain, Dr. Avinash J. Agrawal

Abstract

Twitter and other micro-blogging websites have become an epitome for analysis for user sentiments. In order to perform effective sentiment analysis, researchers need to analyze the data published by users, and then apply different language processing models on the same. A large number of tweets these days are based on political scenarios, wherein twitterati engage themselves in debates over the political decisions, and their impact on daily user’s life. Various methods have been proposed in recent years, and each of them has their distinctive advantages and drawbacks. In this text, we reviewed the state-of-the art methods for sentiment analysis, and tried to keep the work for the political domain. This is done because many cases have come up in recent times, where political campaigns are run by checking out the general user’s sentiments. For example, if a particular party is not-strong in a particular area, then the sentiment analyzer can help in identifying areas of improvement based on the tweets/social media data analysis for that party. Thereby, increasing the chances of winning the elections in that particular area. In this paper, we also recommend some improvements in the existing work, via which researchers can design better sentiment analysis systems.

Keywords: Politics, tweets, social, media, sentiment

Downloads

Published

2020-05-18

Issue

Section

Articles