ENVIRA
From EuroVR Knowledge Base
- Full name
- Intelligent Virtual Learning Environments
- Acronym
- ENVIRA
- Website
- http://www.diana.uma.es/researchIHM/envira.htm
- Keywords
- Virtual Environments; Interaction; Virtual Pedagogical Agent; Case Based Reasoning; Presence; Intelligent tutoring.
- Budget
- € (UPM); 60.500 € (UMA); UCM
- Start date
- October 2006
- End date
- September 2009
Contents |
Partners
Funding Bodies
Contact
Objectives
The general goal of this project is the study of techniques for the construction of effective educational systems based on virtual environments that seek the immersion of the student into a three-dimensional environment, ranging from the conventional desktop systems to more sophisticated virtual reality installations. Regarding the kind of educational approach, we are interested in systems that promote constructive learning involving the student in problem resolution activities, ranging from simulations that reproduce the experience in the real world, to systems that do not intend to reproduce reality but to enhance the learning experience. The concrete goals of this project are aligned with two fundamental work areas: the development of pedagogical virtual agents that assist the student in their learning process, and the study of the most effective interaction mechanisms for each type of application.
With respect to Pedagogical Virtual Agents (PVA), the main hypothesis is that their effectiveness is influenced by their capability to be perceived by the student as intelligent and believable beings. It will be necessary to pay attention to two main aspects: the management of the virtual body and the process of tutoring decision making. Our hypothesis is that believability depends strongly on the capability of the agent to perceive the surrounding environment and particularly the student, as well as on their actuation capability (concentrating on navigation and interaction with objects). We will work on the evolution of an existing visual and auditory perception model for virtual agents that will be extended and complemented with an attention model. Additionally, we will study the application of explicit knowledge representation techniques, namely ontologies, for the construction of semantic descriptions of the environments that will allow the improvement of the PVA actuation capabilities. As for the tutoring capabilities of the PVA, we will take advantage of artificial intelligence techniques, integrating the previous efforts of two of the participating teams. We start from an agent based architecture for intelligent tutoring environments based on virtual reality (MAEVIF), that allows the dynamic computation of solutions for the exercises posed to the student using task planning techniques. On the other hand, we integrate a case based teaching approach where the pedagogical scenarios are the result of the human expert authoring.
As far as the user-system interaction is concerned, this project seeks to deal with a problem for which no universally accepted solutions have already been found and especially interesting in the area of training simulations using virtual reality: the interaction between the user (student) and the virtual environment. In this project we will face two of its possible dimensions: navigation and object manipulation. Consequently, we include two main goals in the project: Development and evaluation of navigation, selection and manipulation mechanisms for virtual environments; and increase of the existing knowledge about the effectiveness of interaction mechanisms in virtual environments, and particularly in training applications
Main Results
End Users
Evaluation Methodologies
Publications
Application & Research Areas
Technical Areas
Comments
Fill with comments.
