A Computational Environment to Experiment with Artificial Intelligence Techniques in Digital Games
"Although one of the fundamental goals of AI is to understand and develop intelligent systems that have all of the capabilities of humans, there is little active research directly pursuing that goal. We propose that AI for interactive computer games is an emerging application area in which this goal of human-level AI can successfully be pursued" [4].
The scope of this project is classical and non-classical logics (multimodal and intensional logics), multi-agent systems (reasoning about knowledge, uncertainty and action - [2, 7, 3, 5]), declarative programming (constraint and intensional programming), distributed computing and games. Logical formalization of agent behaviour is desirable to provide the foundation for sophisticated reasoning techniques to be used on, and by, the agents themselves. In particular, the intensional logics [1, 8] - that include apparatus for signifying when two meanings are identical, and that analyzes inferences involving meanings, of practical reasoning are attractive for their capability of encapsulating practical reasoning. Also, the possible worlds semantics offered by modal logic [3, 7], has proved to be a successful framework in which to model mental attitudes of agents such as beliefs, desires and intentions [6]. More precisely, we examine the relationship between agent (intensional agent) specifications and operational behaviour of multi-agent systems (adapted to different applications scenarios).
Bibliography
- [ 1 ] Benthem, Johan van. A Manual of Intensional Logic, University of Chicago Press, 1985.
- [ 2 ] Fagin, R., Halpern, J. Y. Moses, Y., and Vardi, M. Y. Reasoning About Knowledge. MIT Press, 1995.
- [ 3 ] Halpern, J. Y. Reasoning About Uncertainty. MIT Press, 2003.
- [ 4 ] Laird, J. E., and van Lent, M. Human-level AI's Killer Application:Interactive Computer Games. AI Magazine, 22 No. 2, 2001, 15-26.
- [ 5 ] Moore, R. C. Logic and Representation. CSLI Lecture Notes, No. 39, 1995.
- [ 6 ] Rao, A. S., Georgeff, M. P. BDI Agents: From Theory to Practice. Proceedings of the First Intl. Conference on Multiagent Systems, 1995.
- [ 7 ] Yamamoto, F. S. Sistemas Modais de Conhecimento. Dissertação de Mestrado, IME-USP, 2003.
- [ 8 ] Zalta, Edward N. Intensional Logic and the Metaphysics of Intensionality. MIT Press, 1988.
Towards a Game Architecture for Artificial Intelligence Learning
Interactive computer games can be a powerful tool to be used on Computer Science Education. Researchers are aware of this fact as papers on this subject have appeared in important conferences on education [1, 2, 3]. We believe that interactive computer games are a compelling and effective tool that can improve the way we teach Articial Intelligence in Computer Science and related courses. Our goal in this project is to build a virtual environment in the form of a computer game, which will provide for the students a way of visualizing their programming assignments (agents) in a course of Artificial Intelligence. Given the fact that the virtual environment where the game takes place can be arbitrary complex, we believe that implement inteligent agents that live in such environment becomes an interesting and intricate problem for the students to solve as the course unfolds.
Bibliography
- [1] Randolph M. Jones. Design and implementation of computer games: a capstone course for undergraduate computer science education. In SIGCSE '00: Proceedings of the thirty-first SIGCSE technical symposium on Computer science education, pages 260-264, New York, NY, USA, 2000. ACM Press.
- [2] Ron Coleman, Mary Krembs, Alan Labouseur, and Jim Weir. Game design and programming concentration within the computer science curriculum. SIGCSE Bull., 37(1):545-550, 2005.
- [3] Lasse Natvig and Steinar Line. Age of computers: game-based teaching of computer fundamentals. In ITiCSE '04: Proceedings of the 9th annual SIGCSE conference on Innovation and technology in computer science education, pages 107-111, New York, NY, USA, 2004. ACM Press.
A Truth Maintenance System Implementation
An intelligent agent acts based on its perceptions and on its knowledge base. However, some perceptions makes it clear that the knowledge base itself must be an object passive of action. This happens when a perception enters in conflict with the predicates at the knowledge base. Truth maintenance systems (TMS) are computational efficient mechanisms that aim to make it possible for intelligent agents to revise and update its knowledge base. The final goal of the project is to implement a TMS in a bot of the game Guntactyx and to conclude whether or not this bot wins more battles than bots without a TMS.
Software Engineering for Computer Games
We're currently structuring a software engineering course in which the students will be divided into groups of five or six people and each group will build a simple 2D game engine and make two different games with it. We are discussing and "testing" this project to make sure it is big and complex enough to be done by one group in a semester. We expect the students to organize themselves and put into practice the principles of software engineering while they learn them. A detailed especification about this project, a description of the game engine and what kind of games we can build with it, will be avaible soon.