Predicting the Popularity of Political Parties through Ensemble Learning

International Journal of Research and Scientific Innovation (IJRSI) | Volume VI, Issue V, May 2019 | ISSN 2321–2705

Predicting the Popularity of Political Parties through Ensemble Learning

Teenu Sharma¹, Ankita Bhargava², Shruti Jain³

^1,2Department of Computer Science and Engineering, Pacific Institute of Technology, Udaipur, Rajasthan, India
³Department of Computer Science and Engineering, Indore Institute of Science and Technology, Indore, Madhya Pradesh, India

Abstract— With the advancement in Technology, social media has become a part of our daily life. People use it to share their day-to-day activities, likes, dislikes, opinions regarding any product, service or event. The micro-blogging website Twitter is a rich source of opinionated content where almost 500 million tweets are sent every single day. This rich opinionated content can be used for analysis, studies, research and it can provide beneficial results. In this paper, tweets are extracted from Twitter for the upcoming India General Elections 2019 and Sentiment Analysis (SA) is performed on it. Three classification algorithms- Naive Bayes (NB), Support Vector Machine (SVM) and k-Nearest Neighbor (k-NN) are used to assign polarity to tweets as positive or negative. Then, the accuracy of predictions is improved through Ensemble Learning and based on this, the popularities of both the parties is calculated and compared.

Keywords— Classification, Classifier Combination, Ensemble Learning, Hard Voting, Majority Voting, Multiple Classifier System, Natural Language Processing (NLP), Sentiment Analysis

I. INTRODUCTION

Now-a-days everything is becoming digital. Life is almost nothing without data. Data is growing exponentially. Each and every domain of life is directly or indirectly connected to data and on the information processed from it. With the advances in technology, it has become very easy to generate, collect and process data. This data is a blessing for Machine Learning.

Machine Learning is an application area of Artificial Intelligence (AI) which focuses on the development of computer programs (i.e. machines) in such a way that they automatically access the data, use it to learn by themselves and then, process real-time data on their own without the intervention of humans.

One of the most trending fields for machine learning, with enormous amount of data is Social Media. Social Network Platforms like Facebook, Twitter, Instagram etc. have gained popularity because they are easy modes of communication and information. The most popular platform is Twitter where people communicate with each other by posting tweets, retweeting them, liking and commenting on tweets, uploading pictures and sharing interesting videos. One of the most interesting aspects of Twitter is that people express their sentiments in their tweets and comments.

Read More