Home // International Journal On Advances in Software, volume 14, numbers 1 and 2, 2021 // View article

Methodology for Extracting Knowledge from a Gaming Simulation Using Data Envelopment Analysis

Authors:
Akinobu Sakata
Takamasa Kikuchi
Ryuichi Okumura
Masaaki Kunigami
Atsushi Yoshikawa
Masayuki Yamamura
Takao Terano

Keywords: gaming simulation; data envelopment analysis; knowledge extraction; debriefing; facilitation

Abstract:
In this study, we propose a methodology for extracting knowledge about which play logs are superior (inferior) to other play logs under certain criteria from the results of a gaming simulation. Previous research has enabled facilitators to know where players’ play logs output from gaming simulations are positioned in all possible scenarios. However, facilitators have no valid solution to encourage players to change their behavior in gaming simulations. The proposed methodology enables a facilitator to identify the players who show similar behavior and performance to the target player under certain criteria, and to present to the play logs which show superior performance than the target player’s play log to the target player. In order to achieve our research objective, we created several software agents instead of human players to play a gaming simulation for career education, and analyzed the output play logs using data envelopment analysis. As a result, the desired knowledge was extracted. We argue that the extracted knowledge should be applied for debriefing. The proposed methodology is flexible enough to work under both conditions where all players are human and where human and machine agents are mixed as players.

Pages: 107 to 121

Copyright: Copyright (c) to authors, 2021. Used with permission.

Publication date: December 31, 2021

Published in: journal

ISSN: 1942-2628