LudoScience

Using Players? Gameplay Action-Decision Profiles to Prescribe Training: Reducing Training Costs with Serious Games Analytics Christian Sebastian Loh, I-Hung Li - 2016

Informations

Support : Références scientifiques
Auteur(s) : Christian Sebastian Loh, I-Hung Li
Editeur : Data Science and Advanced Analytics (DSAA), 2016 IEEE International Conference on
Date : 2016
Langue : Langue


Description

Abstract— Players’ gameplay action-decision data can be used towards profiling as serious games analytics. The insights gained can help support the decisions for performance improvement and as ‘prescriptions’ for training – e.g., diagnosing who should receive training, how much training will be given, informing the design of the game, and determining the contents for inclusion and exclusion. Data-driven training prescription can help learning organizations save money by mitigating unnecessary training to reduce costs. Players’ learning performance in games can be measured in lieu of their behaviors traced in situ the training environment. Novice players’ actiondecision data can first be converted into Course of Actions (COAs) before pairwise similarity comparison against that of the expert(s) to reveal how similar they are to the training goal, or expert/model answer. We identified three Gameplay ActionDecision (GAD) profiles from these gameplay action-decision data and applied them as diagnostics towards prescriptive training. 

 

Références (1) :

 

J. Alvarez, D. Djaouti, O. Rampnoux, and V. Alvarez, “Serious games market: Some key figures (from 1950’s to 2000’s),” 2011. [Online]. Available: http://serious.gameclassification.com/files/articles/sgc_report_ 03-11.pdf. 



Mots-clés : Serious Games Analytics, Similarity measures, Performance improvement, Reducing training cost, Training prescription