Intelligent Learning Environments

Report from the project "Les systèmes explorateurs intelligents"

P. Dillenbourg, M. Hilario, P. Mendelsohn, D. Schneider, B. Borcic

Abstract

Table of Contents

1. Goals and Products

2. Research Methodology

3. Overview of MEMOLAB
3.1. Quality #1: The learner must be active.
3.2. Quality #2: The environment must be rich and complex
3.3. Quality #3: The environment must be structured
3.4. Quality #4: The learner must interact with agents

4. Agents in ETOILE
4.1. Role distribution
4.2. How to implement the interaction between the learner and a computational agent?
4.3. Diagnosis or Agreement
4.4. The learner-expert-tutor triangle
4.5. Teaching styles
4.6. Selecting teaching styles

5. Developing a new learning environment with ETOILE

6. Implementation

7. Experiments

8. Synthesis & Future Developments

Appendix I:
The Simulation
I.1. Pedagogical constraints.
I.2. Knowledge Acquisition
I.3. The Case Library
I.4. The Algorithm
I.5. Output from the simulation

Appendix II:
The Integrated Hypertext
II.1. Design
II.2. Implementation

Appendix III:
The Inference Engine
III.1. The OOPS system
· tight integration of object-oriented and rule-based programming
· possibility of concurrent multi-expert reasoning
· full computational support of metareasoning
III.2. Main principles of the inference engine
III.2.1. Deliberation cycle
III.2.2. Action cycle
III.3. Adapting OOPS to ETOILE
III.3.1. Covering all solutions: optimal rules, suboptimal rules and repair rules
III.3.2. Diagnosis: rule preview
III.3.3. The integration of procedures
III.3.4. Control among agents

Appendix IV:
The Design Metaphors
IV.1. The metaphors
IV.2. Theoretical background
IV.3. Applying design metaphors to MEMOLAB

9. Bibliography