REflective Agent Learning environment project (REAL)

About REAL

REAL is an agent-based educational gaming environment that allows students to (1) construct an imaginary world; (2) reflect upon the quality of their understanding of that world; and (3) test this understanding out in a dynamically generated simulation game.

The critical component of the REAL framework is the stress on reflection as part of the thinking process. The use of the term “reflective” gives the agent two layers of meaning: on the one hand, it reflects the users’ domain knowledge back to them (similar to looking into a glass box and seeing their own reasoning); on the other hand, the users reflect on that knowledge (similar to looking at a mirror, criticizing their own thinking processes). In developing REAL, we are exploring ways to design a framework that allows users to expose their thoughts and pass them to a reflective agent, and then to let the reflective agent represent them in making decisions and taking action in a simulated gaming environment. The reflective agent’s action is monitored by an expert agent, a pedagogical agent, and a communication agent, using an embedded rule-based reasoning and gaming engines. REAL will make it easier for developers to model domain knowledge and develop simulation games, which are otherwise time-consuming processes.

Preliminary studies have shown that this kind of learning environment not only engages students in learning but encourages collaborations among researchers in subject domains.

Figure 1: REAL learning environments - a meaningful connection to the real world

The REAL Design

The REAL framework benefits from existing intelligent computing approaches, such as ITS, ICAI, and Expert Systems. It is designed as a framework to inform software design that will support four kinds of globally used knowledge representation. It uses a rule-based engine for reasoning, based upon declarative/procedural knowledge, and a game engine for representing mental images and imaginary worlds.

Figure 2: The REAL cognitive architecture

Figure 3: REAL’s reusable system architecture

The REAL Implementation

REAL Planet and REAL Business are two prototype implementations that use the REAL framework.

REAL Planet

REAL Planet was developed in the domain of ecology. In order to generate their own simulation games, students teach an alien how to design an ecological system on an alien planet that has environmental conditions similar to that of Earth. The agents in an ecosystem interact locally with each other and with their environments. The global behavior of the entire ecosystem cannot be predicted beforehand--it is an emergent outcome of the interactions among the agents.

We have designed the agents' properties at two levels:

(1) The cognitive level: Agents are based upon definitions of reasoning and behavior, including such properties as predator-prey (energy flow) relations, categories (herbivore, carnival, producer or decomposer), goals, and behaviors.

(2) The biological level: Properties in this level are controlled by a thread built into each agent that is .adjusted in each bio-clock tick and includes such properties as energy level, age, growth rate, reproduction rate, etc. Each agent will perceive what is happening in the environment (sense), make decisions (reason), and change the environment by interacting with other agents (act).

Figure 4: Design Modes in REAL Business

REAL Business

REAL Business was developed in the domain of probability. Students help a store owner design business strategies to run a successful ice-cream store. In Design Mode, students are provided with tools to view entities (such as possible events), as well as relations between entities (such as the probability of an event happening after a certain condition arises). Students complete procedural rules by reasoning about the likelihood of possible events happening. These facts and rules guide the reflective agent’s behavior in Game Mode, a simulated real world with multiple game levels. Students use the Game Mode to analyze the performance of their agent, i.e., whether the rules they have created work. Students can also use the Reflection Mode to refer to real-time store reports to inspect the results of their self-generated procedural rules.

Figure 5: REAL Business Design Modes

Figure 6: REAL Business Game Mode and Reflection Mode

Publications:

Bai, X., Black, J. B., Vitale, J. (2007). REAL: An agent-based learning environment. Paper presented at the Agent-Based Systems for Human Learning Conference, Hawaii. Word

Bai, X., Black, J.B., Vikaros L., Vitale, J, Li, D., Xia, Q., Kim, S., Kang, S. (2007). Learning in One’s Own Imaginary World. Paper presented at the Annual Meetings of the American Educational Research Association, Chicago. PDF PPT

Bai, X., & Black, J. B. (2006). REAL: An agent-based learning environment. Paper presented at the Agent-Based Systems for Human Learning Conference, Hakodate, Japan.

Bai, X. (2006). Modeling Users Through Self-Reflection in Simulation Games. Paper presented at the Association for Psychological Science. 18th Annual Convention, New York, NY.

Bai, X., & Black, J. B. (2005). REAL: A Generic Intelligent Tutoring System Framework. In C. Crawford et al. (Eds.), Proceedings of Society for Information Technology and Teacher Education International Conference 2005 (pp. 1279-1283). Chesapeake, AACE.

Bai, X., Saravanos, A. & Black, J. (2005). REAL: Facilitate Thinking in an Object-oriented Way. In Proceedings of World Conference on Educational Multimedia, Hypermedia and Telecommunications 2005 (pp. 3383-3387). Chesapeake, VA: AACE.

Bai, X., & Black, J. (2004). TALE: A Teachable Agent Embedded in an Intelligent Tutoring System. In G. Richards (Ed.), Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education 2004 (pp. 1070-1072). Chesapeake, VA: AACE.