Charlie Waddell Undergraduate Dissertation 2016/17

Event Detection for Sports using Twitter

Supervised by A.Vlachos

Abstract

The evolution of Twitter has sparked an increase in the research of event detection using social media. With the microblogging service receiving over 500 million tweets per day, information about events such as sports becomes available to us in real-time – an important notation consistently referred to throughout. This project aims to develop an intelligent system that uses tweets to detect events that occur within football matches. This is explored through the domain of Machine Learning, where a distantly supervised approach has been followed along with the Multinomial Naïve Bayes algorithm to successfully detect events.