Popularity Measuring and Prediction Mining of IPL Team Using Machine Learning

Abstract

The Indian Premier League (IPL) is a Twenty20 cricket league held in India each year between March and May. Eight teams participating from Indian cities or states. The Board of Control in India (BCCI) started the league in 2007. The Indian Premier League (IPL) is the most popular cricket league in the world, ranking sixth among all sports leagues in terms of average appearance in 2014. The IPL was the first sporting event in the world to be broadcast live on YouTube in 2010. Since the IPL is so common, In order to be more competitive, it is necessary to assess how people will react to the league. Machine Learning techniques can be used to resolve the problem of predicting outcomes. Public opinion, as well as the causes behind it, has been discovered to be political. Among the many social media channels available, one of the popular social service websites is Twitter, which allows users to post events from their daily lives. 7 lakh People's feedbacks are taken from the IPL-2020 Twitter website using Application Interface. This paper is split into two sections. First, we divided 7 lakh tweets into different teams based on hashtag attributes, analyzed people's feelings in neutral, positive, and negative, checked the most popular team during IPL-2020, and measured accuracy using Multinomial Naïve Bayes, Logistic Regression, Ridge Classifier, Random Forest, SGD Classifier, and Decision Tree on those tweets. In the second part, 50k data were collected at random from 7 lakh tweets and F1-score, precision, accuracy, and recall was measured using Multinomial Naïve Bayes, Logistic Regression, Ridge Classifier, Random Forest, SGD Classifier, and Decision Tree. This paper aims to correctly assess people's feedback for an event such as the Indian Premier League 2020.