Towards principles of sound combination in audio games using machine learning

Date of Publication


Document Type

Master's Thesis

Degree Name

Master of Science in Computer Science


College of Computer Studies


Computer Science

Thesis Adviser

Raymund C. Sison

Defense Panel Chair

Rafael A. Cabredo

Defense Panel Member

Raymund C. Sison
Solomon L. See
Merlin Teodosia C. Suarez


Audio is one of the most important element s in every game. In recent years, audio and exercise games have been gaining in popularity since the release of the Microsoft Kinect and Wii. Nevertheless, there is still little research done on this aspect of games especially in the field of sound combination. This research paper used an audio exercise game and machine learning to be able to come up with sound combination principles based on two types of evaluation metric; performance and fun. The research used sounds belonging to the Zone and Affect category of the IEZA Framework. Also Sweetsers GameFlow model was used for the evaluation of the fun metric. The research discovered five principles of sound combination which can be used when designing audio games specifically the boxing game genre. Furthermore, the research also discovered a certain threshold in the number of sounds present before a decrease in performance and overall fun was observed. Lastly, the research was able to categorize the participants based on their playstyle in relation to Bartles player types..

Abstract Format






Accession Number


Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Physical Description

1 computer disc ; 4 3/4 in.


Machine learning; Video games; Electronic games; Computer games

This document is currently not available here.