Document Type

Student Research Paper

Date

Spring 2020

Academic Department

Computer Science

Faculty Advisor(s)

Peilong Li

Abstract

Sentiment analysis is a topic in natural language processing that seeks to automatically extract positive and negative polarity from text data. Its applications are diverse, ranging from marketing and sales to forum moderation to gauging public opinion. One particularly interesting application area is found in professional sports: fans share a huge volume of opinions, predictions, and reactions online that can be used to monitor public opinion on specific teams, coaches, and players. This paper explores the application of machine learning based sentiment analysis on a hand-labeled social media dataset focused on reacting to National Football League draft picks. The resulting model, called DraftSense, provides information that can be used for future analysis, including attitude towards drafted players, comparison between fan reactions and on-field performance, and comparison between drafted players based on the language used to describe them. Additionally, a labeled dataset for sentiment analysis on professional football will be created for further use.

Notes

Senior thesis.

Download

Share

COinS