DIGITAL LIBRARY
EVOLVING NON-PLAYER CHARACTERS IN EDUCATIONAL GAMES IN VIRTUAL WORLDS
University of Patras (GREECE)
About this paper:
Appears in: EDULEARN23 Proceedings
Publication year: 2023
Pages: 7053-7059
ISBN: 978-84-09-52151-7
ISSN: 2340-1117
doi: 10.21125/edulearn.2023.1851
Conference name: 15th International Conference on Education and New Learning Technologies
Dates: 3-5 July, 2023
Location: Palma, Spain
Abstract:
3D virtual worlds (VWs) are an integral part of the landscape of education and demonstrate new learning practices in education. Non Player Characters (NPCs), the embodied agents, that support the user, constitute a vital part of 3D VWs. The impact of users’ interaction with NPCs and especially with the Pedagogical Agents (PAs), the ones that support the pedagogical process can be positive. The support of PAs can be adapted according to users’ profile and behavior.
In the context of Erasmus+ projects, we have developed educational games in VWs (AVARES, Agrient, Biz4Fun, NET). We have implemented various NPCs according to the needs of each environment and the project objectives. Their role was to introduce the Virtual World environment, inform the users about the educational game, the learning and assessment activities and accompany them offering tips. However, they did not have real interaction with the user. A study using VARAES VW game showed that the presence and the behavior of the PAs can have substantial impact on student’s learning, increase their engagement and improve their learning experience.

In this work, we present how we plan to extend our VW game architecture approach to include logging analysis model. In the game, various activities will be implemented, and different types of PAs will be involved in the activities. The logging information in the log files is:
• time spent in the game / in each module /in each activity
• participation in the activities (time pent, percentage of completion, return to activity, sequence of activities)
• interaction with resources (presentations, tests, multimedia resources),
• interaction with PAs (time, PA involved, type of interaction: question and answer/feedback/tip/ personalized feedback, result of interaction: guided to relevant action/ ignored/not applicable)
• score (activity/game)
• sum of interventions of PA with each learner / in total.

Learner’s log files will give us, during the game, an insight view of his/her activity through his/her individual history and actions. These analytics will be used for the PAs to offer adaptive support to the learners. A study with learners for the evaluation of the proposed approach will be also presented.
Keywords:
Virtual 3D worlds, NPC, e-learning.