Automatically Generating Summary Visualizations from Game Logs

In this paper we describe a system called ViGLS (Visualization of Game Log Summaries) that generates summaries of gameplay sessions from game logs. ViGLS automatically produces visualization of the summarized actions that are extracted based on cognitive models of summarization. ViGLS is implemented using a serviceoriented architecture, de-coupling the summarization methods from any particular game engine being used. The camera code libraries used in visualization are based on constraint based camera control approaches and, in our implementation, make use of the scripting layer of the Unreal game engine.

• Type: Conference
• Year: 2008
• Where: AIIDE
• Authors: Yun-Gyung Cheong, Arnav Jhala, Byung-Chull Bae, R. Michael Young
• Keywords: AIIDE 2008, Artificial Intelligence, Game Engine, Game Log Summaries, Game Logs