In 2019, I undertook an Extended Project (EPQ) to produce a dissertation-style report to investigate the following project title. Can Machine Learning Harness the “Wisdom of the Crowd” to Predict the Outcome of Football Matches? I chose the topic because of my strong interest in machine learning and natural language processing; I fused this with my passion for football. I started by an thorough literature review to understand the state-of-the-art. Subsequently, I decided to design and build my own attempt at Twitter sentiment analysis football prediction. This involved developing a data pipeline to pass filtered Tweets to a sentiment analysis model I made; the polarity (positivity) of the Tweets from a fanbase were collated and passed through a linear regression model in TensorFlow. My model was able to effectively identify events in matches such as goals scored/conceded. Although, the model was not an effective outcome prediction model, which was expected with points raised earlier in the report that social media 'Wisdom of the Crowds' will not provide a standalone model.
-
Notifications
You must be signed in to change notification settings - Fork 0
liamhbyrne/twitter-football-prediction
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Sentiment analysis on football fans to correlate a relationship between polarity and results. Use of Sentiment Analysis and Linear Regression to enable "wisdom of the people" approach.
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published