|Abstract || Project Summary/Brief Description of the Infrastructure
In language appropriate for a review committee, use the space below and up to one additional page to provide a general description of the infrastructure requested and indicate where it will be located. Briefly explain why the infrastructure is needed at this time, how it enhances any existing infrastructure, what research or technology development it will enable to be performed, and why that research or technology development is important for Canada.
With the development of virtual worlds and game platforms, such as Wonderland, World of Warcraft, Second Life, Open Croquet, and There.com, many universities have been engaged in developing a range of virtual environments for distance education (Coffman & Klinger, 2007; Callaghan et al., 2008).
Athabasca University (AU) is an international leader in online and distance education research and practice. AU researchers are actively investigating interactive virtual learning environments and improving their usability, applicability, and effectiveness using technological innovations in multi-modal affective tele-interaction.
This proposed infrastructure will be used to conduct research to develop a framework and software architecture for developing effective immersive game-based learning environments with motivational and immersive traits. An immersive game-based learning environment is typically a dynamic three-dimensional (3D) “learning space” that is constituted by the context, the user, the content, the learning objects, and the learning activities. We will propose a novel framework for such environments through incorporating the adaptivity and intelligence features into the environments to enhance the usefulness and usability. The proposed framework will integrate a multi-agent system (MAS) into a game-based virtual world by modeling the dependent, human-computer components in such a space. We aim to inquire the adaptive, collaborative, and interactive possibilities and beyond among the human-computer components in the 3D environment through addressing the latest development and application of human-computer interaction (HCI) in the 3D virtual learning environments.
While MAS providing much necessary promising functionality such as control and coordination, intelligence, and autonomy, for adaptive and intelligent 3D virtual learning environments, their effects in human learning are still in progress. A variety of technical challenges, such as differences in concern between agent technology and game technology with regard to efficiency, distributed control, and level of autonomy, and pedagogical challenges such as learning strategies, successful factors in promoting learning, learning activities design, learning preferences/styles modeling, needs to be addressed. The questions to be answered in the proposed project are:
We aim to address these issues through selecting a multi-agent platform JADE (Bellifemine et al., 2005) and a game engine Sun’s Project Wonderland (WL, for short). Although we aim to investigate solutions to specific aspects of 3D virtual environments (3DVE) educational game development using WL, through the implementation of a working prototype, we will prove that our approach can bring essential educational components to 3DVE to produce effective and engaging immersive educational learning activities. We will test and verify our solution though a series of scenario-based functional tests.
We will use agent technology to maintain user models and interactive learning activity elements, which allows us to incorporate intelligence and adaptivity features into the games, making these characters more believable, and adding to the overall fantasy of educational games. An important outcome of our research will be to identify a suitable mechanism for the exchange learning activity information between the WL 3D environment and the JADE muiltiagent system. We propose a messaging system in which ‘percept’ messages are sent from WL to the MAS, and ‘action’ messages are sent from the MAS to the WL environment. Using this approach, the MAS system should be able to detect and react to important game events in the WL environment. In order to test and verify our approach, we will implement a prototype multi-player educational game based on a TV game-show theme. The research will provide the educational game community with a framework on which to base further research of 3DVE game development using the WL toolkit.