Vote Bank Forecast Based on Social Media Sentiment Analysis

Sentiment Analysis of human behavior from social media revolutionized many fields, such as health, business, and election prediction. Sentiment analysis is the process of analyzing an individual’s opinions, attitudes, views, sentiments, and emotions. In order to predict sentiments, predicting the polarity of words and categorizing them as positive or negative are the major goals. A quick and effective way to determine what the public thinks about one’s own political party, the opposition party, and political figures is through Twitter. prior sentiment analysis techniques have primarily struggled with classification accuracy because biases in the training data cause a sizable portion of tweets to be incorrectly classified. As a remedy, we set out to collect real-time data in order to forecast Pakistan’s next election results. Users can quickly share real-time updates on a variety of activities and events on Twitter, promoting the rapid diffusion of information and news. To do this, we accessed the Twitter API through the “Tweepy’’ Python package and retrieved real-time data from the service. Through keyword searches, we were able to find more than 5,000 tweets. We attributed feelings to particular tweets using Python’s “TextBlob’’ package. We are able to generate prediction graphics by determining the frequency of positive attitudes with respect to a particular candidate.