Automating Content Analysis of Video Games

Authors

  • Terry Bullen
  • Mike Katchabaw
  • Nick Dyer-Witheford

Abstract

Content analysis of video games tends to be an extremely arduous task, involving the collection of a very large quantity of data and statistics detailing the experiences of gameplay. Nevertheless, it is an important process that supports many business, policy, social, and scholarly activities related to the games industry. Consequently, supports are clearly necessary to facilitate content analysis procedures for video games. This paper discusses an innovative approach to automating content analysis for video games through the use of software instrumentation. By properly instrumenting video game software to enable data collection and processing, content analysis procedures can be either partially or fully automated, depending on the game in question. This paper discusses our overall approach to automation, as well as our experiences to date with Epic’s Unreal Engine. Sample results from initial experiments conducted so far are also presented. These results have been quite positive, demonstrating great promise for continued work in this area.

Downloads

Issue

Section

Foundations and Frames