Web application with spatio-temporal analytics on soccer game data

Abstract

As previous methods to process data collected from sports games and retrieve similar plays are too inefficient and ineffective, a new method called play2vec was devised to improve on past methods and tackle their shortcomings. In this Final Year Project, a web application that integrates play2vec into its core logic is created for users to retrieve similar plays within the same dataset provided.

The paper describing play2vec was reviewed to understand the algorithms and methods used so that they can be integrated into the application. The web application architecture is designed based on a front-end back-end architecture for easier maintenance and development. Since play2vec was written in Python, the back-end consists of a Flask application that uses play2vec functions, whereas ReactJS was used to develop the front-end user interface. The requirements were drafted before the development of the application, of which is demonstrated as images in this report.

Finally, the report is concluded with a description of how the application integrates play2vec into its core logic, shortcomings of the application and possible future workings on the application.