Real-Time Sentiment Analysis of Tweets: A Case Study of Punjab Elections
Yachika Gupta, Parteek Kumar · 2019
Twitter, a micro-blogging site, is a huge repository of public opinions expressed towards various people, services, organizations, products etc. Sentiment analysis is the process of analyzing those public opinions. Sentiment analysis when combined with twitter gives useful insights into what is expressed on Twitter. It has applications in a number of domains like in stock market prediction, election results prediction, movie revenues, product reviews etc. This paper discusses about the different machine learning and deep learning models trained on a dataset of tweets collected from online github directory. The proposed system has been validated using a case study of election results prediction of February 2017 Punjab elections by analyzing public sentiments as expressed on twitter towards various political leaders. The developed SA system performs real-time sentiment analysis presenting analysis results simultaneously as tweets occur on twitter. A dashboard has been designed for presenting results to the user. The display updates every one minute, bringing tweets in real time and presenting the results graphically. Tweets are also getting stored in a CSV file for final results prediction. The system gave very encouraging results with minimal diversion when compared with actual results.