Playphysics: an emotional student model for game-based learning

Abstract

Game-based learning (GBL) environments introduced a new generation of Intelligent TutoringSystems (ITSs) that provide personalised instruction by being constantly aware of studentreactions to the system. Student motivation, attitudes, self-efficacy and affective statehave been the key focus of such developments. Current models of student emotion haveshown promise in laboratory environments. However, the problem of accurately recognisingand inferring student emotions within learning environments persists. The majority of alreadyexisting computational models of student emotion employ cognitive theories that are not derivedfrom the learning context.Control-value theory (Pekrun et al. 2007) assumes that control and value appraisals arethe most meaningful for determining emotions in educational settings. Our proposed computationalemotional student model uses the Control-value theory for reasoning about learners’emotions in GBL environments settings. The main hypothesis is that this model will recognisestudent achievement emotions, i.e. emotions relevant to the educational context, withreasonable accuracy (not random). The definition, implementation and evaluation of ourcomputational emotional student model in PlayPhysics, an emotional game-based learningenvironment for teaching physics, are discussed. Our emotional model is implemented with adynamic sequence of Bayesian networks for representation of learners’ achievement emotions.Probabilistic Relational Models (PRMs) are employed to facilitate their derivation. TheNecessary Path Condition algorithm is employed in combination with Pearson correlationsand Binary and Multinomial logistic regression for defining network structure. The ExpectationMaximisation (EM) learning algorithm is employed for network parameter learning. Ourmodel employs answers to questions in-game dialogues, contextual variables and physiologicalvariables for recognising student emotion.Results show a fair accuracy of classification of student achievement emotions for thePlayPhysics’ emotional student model when only contextual and behaviour variables areconsidered (values of Cohen’s Kappa in a range larger than 0.2 but lower than or equal to0.4), which then improves when physiological variables, i.e. Galvanic Skin Response (GSR),are incorporated (values of Cohen’s Kappa in a range larger than 0.4 and lower than or equalto 0.6). Our emotional model provides enhanced understanding about the factors involved inreasoning about emotion. PlayPhysics GBL environment is assessed to attain an enhancedunderstanding of the student experience of achievement emotions. Future work may focuson creating further game challenges, identifying enhanced predictors for control and value,e.g. using sentiment analysis and analysis of facial expressions. Numerous applications, inareas ranging from biology to e-commerce, are envisioned for the application of our approachto create intelligible and dynamic genetic and emotional consumer data models.

Date of Award	Jul 2012
Original language	English
Awarding Institution	Ulster Univeristy
Supervisor	Paul Mc Kevitt (Supervisor) & Tom Lunney (Supervisor)