Abstract:

Serious games have been increasingly used in the fields of education and assessment, especially for the stealth assessment of competences. Previous efforts have been broadly designed for serious games in general, without specifically focusing on the pedagogical aspects of assessment. In fact, personalization is crucial to guarantee customized game structures and scenarios that fit the learner’s profile, portfolio, and gameplay, in order to achieve authentic assessment results. Under the context of our framework Lebanese-Egyptian-Tunisian Serious Games for Assessment (LET’SeGA), which aims to use serious games for assessment (SGA) to assign students to respective college programs, we proposed in a recent work an ontology that captures game content and its personalization with respect to various players groups defined by similar players profiles, ePortfolios, learning styles, and gameplay styles. In another work, we proposed a set of reasoning rules on top of this ontology to show how in-game personalization can be defined in SGA. In this paper, we use segmentation to define groups of players and we map them to several game adaptations, and we show how some game content types can be generated and auto-adapted with target players groups. In addition, we link games to academic program objectives by expressing which competencies required by a given program are assessed by which set of game scenarios. Finally, we aim to evaluate our proposed personalization methods by applying those techniques to given sets of students and matching the assessment results with those of the pedagogue.

Keywords:

Serious Games, Assessment, Personalization, User Identification, Content Generation.