Predicting Election Results using NLTK

Abstract

In today’s world, people are usually using social media networks for trying to communicate with other users and for sharing information across the world. The online social networking sites have become considerable tools and are providing a common medium for a number of users to communicate with each other. Twitter is the most prominent microblogging website and one among the social networking sites that grow on a daily basis. Social media incorporates an extensive amount of data in the form of tweets, forums, status updates, comments, etc. in an attempt to automatically process and analyze these data, applications can rely on analysis approaches such as sentiment analysis. Twitter sentiment analysis is an application of sentiment analysis on data from Twitter (tweets), to obtain user's opinions and sentiments. Natural Language Toolkit (NLTK) is a library based on machine learning methods in python & sentiment analysis tool. Which provides the base for text processing and classification? The research work proposed a machine learning-based classifier to extract the tweets on elections and analyze the opinion of the tweeples (people who use twitter). The tweets can be categorized as positive, negative and neutral towards a particular politician. We classify these processed tweets using a supervised machine learning classification approach. The classifier used to classify the tweets as positive, negative or neutral is Naive Bayes Classifier. The classifier is trained with tweets bearing a distinctive polarity. The percentage of positive and negative tweets is then measured and graphically represented.