home

=**Welcome!**= Hello. This is made for my research topic **Adaptive Agents for Real-time Strategy Game**. It's under construction now. If you are interested in my research topic and want to contact me, feel free to email me. (you could find my email address in [|my profile])

=Content= Problem Definition Supporting Reading Proposed Solution Scratch Idea People

 =Problem Definition=

(Updated in Sep, 2008) Currently I focus on identify some formation pattern in the game snapshots, which are also extract from the Bos Wars. Some spatial data mining techniques are used. This requires we have a bunch of game play log. At this phrase I play the game to creat some test data. Depend on the result of this research, I wish later on the technique can be used to help reconganizing the opponent's intentions.

Currently, I am working on to export the game play information from Bos Wars(version2.4.1). Trying to gain some action pattern of the units and, if possible, to form a knowledge base for Bos Wars. The mining would relate to event sequence, spatial informations. Also I am considering how to cluster the units to group (for mining group behavior).

Yes, they're tougher than I thought, especially when my supervisors recommend to add the interrelation between the player's own action and his opponent's reaction. At this moment I am trying to find some ideas to combine them in a rational way. Read something about Game Theory, but I could not figure out whether they are suitable to introduce or how to introduce them into the mining / knowledge discovery in the action sequences.

To define the game AI for the Real-Time Strategy game, we could consider following two usage of the AI.
 * //(The following sections are out-updated. Later if possible I wish I could revise them.)//**


 * 1) The AI bot act as an autonomous player of the game, either as an ally or an opponent of the human players.
 * 2) The AI act as an unit in the game, either being controlled by the human players or AI bot. This AI would carry out the command given by the player(human or AI), with accuracy and intelligence.

//Challenges at unit level include accurate path finding and allowing units a degree of autonomy in order to be able to behave sensibly without the player’s direct control. ---from C. Fairclough// //et al., Research Directions for AI in Computer Games//

The following contain some "brain-storm" style of what I thought about the research topic.

Concepts and relations needed to be investigated include: Turn-based Strategy Game and Real-time Strategy Game, Game Strategy Representation, interactiveness of the strategy games...

About the AI model: Rule-based system and how the game strategy relates to the rules; Short-term planning and long-term planning in the AI model; Game theory related to the strategy decision making.

How to model the game world representation in the strategy generator engine?


 * Next Step or Planning to read or do…**
 * The learning technique from cases or from randomized trials
 * Reasoning from the knowledge base
 * Planning technique to organize sequential tasks or actions to achieve a goal and its sub-goals.
 * Typical real-time strategy game's strategy and how these strategies could be decomposed to some rules or other representation which could be process by computer.

Currently, the several papers do not give me any clue on how to diminish the searching space of the strategy. It seems the only proposed solution is to use case-based method. In this way the points in strategy space are actually clustered into some case points or be rounded up to them. I will try to find out more background knowledge about this.

 =Supporting Reading=

**//Dec 2007//**
Michael Buro: **Real-Time Strategy Games: A New AI Research Challenge**. [|IJCAI 2003]: 1534-1535

[|Santiago Ontañón], [|Kinshuk Mishra], [|Neha Sugandh], [|Ashwin Ram]: **Case-Based Planning and Execution for Real-Time Strategy Games.** [|ICCBR 2007]: 164-178

Thomas R. Hinrichs, [|Kenneth D. Forbus]: **Analogical Learning in a Turn-Based Strategy Game.** [|IJCAI 2007]: 853-858

[|Marc J. V. Ponsen], [|Héctor Muñoz-Avila], [|Pieter Spronck], [|David W. Aha]: **Automatically Acquiring Domain Knowledge For Adaptive Game AI Using Evolutionary Learning**. [|AAAI 2005]: 1535-1540

Avi Pfeffer, [|Ya'akov Gal]: **On the Reasoning Patterns of Agents in Games.** [|AAAI 2007]: 102-109

[|Stefan J. Johansson]: **On using multi-agent systems in playing board games.** [|AAMAS 2006]: 569-576

[|Megan Smith], [|Stephen Lee-Urban], Hector Muñoz-Avila: **RETALIATE: Learning Winning Policies in First-Person Shooter Games.** [|AAAI 2007]: 1801-1806

[|Manu Sharma], [|Michael Holmes], [|Juan Carlos Santamaria], [|Arya Irani], [|Charles Lee Isbell Jr.], Ashwin Ram: **Transfer Learning in Real-Time Strategy Games Using Hybrid CBR/RL.** [|IJCAI 2007]: 1041-1046

[|Bikramjit Banerjee], [|Peter Stone]: **General Game Learning Using Knowledge Transfer**. [|IJCAI 2007]: 672-677

Sander Bakkes and Pieter Spronck (2006). **[|Gathering and Utilising Domain Knowledge in Commercial Computer Games]**. (Presented at the [|BNAIC 2006]).

Sánchez Ruiz-Granados, A., Lee-Urban, S. & Muñoz-Avila, H., González Calero, P.A., Díaz Agudo, B., 2007: "**Game AI for a Turn-based Strategy Game with Plan Adaptation and Ontology-based retrieval**". Proceedings of the ICAPS-07 Workshop on Planning in Games. AAAI Press.

**//Jan 2008//**
Zhou Pu-Cheng ; Hong Bing-Rong ; Huang Qing-Cheng ; Javaid Khurshid: **Hybrid Multiagent Reinforcement Learning Approach: The Pursuit Problem**, [|Information Technology Journal], 2006 Volume: 5 - Issue: 6, page 1006-1011

[|Agnar Aamodt], [|Enric Plaza]: **Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches.** [|AI Commun. 7](1): 39-59 (1994)

Ponsen, M.J.V., Spronck P., Tuyls, K., **Hierarchical Reinforcement Learning with Deictic Representation in a Computer Game ,** BNAIC 2006, October 5-6, Namur, Belgium, 2006

Sander Bakkes, Pieter Spronck and Jaap van den Herik, **Phase-dependent Evaluation in RTS games**. Proceedings of the 19th Belgian-Dutch Conference on Artificial Intelligence, pp. 3-10. (2007)

Laurens van der Blom, Sander Bakkes, and Pieter Spronck (2007). **Map-Adaptive Artificial Intelligence for Video Games**. 8th International Conference on Intelligent Games and Simulation (GAME-ON 2007), pp. 53-60

C. Guestrin, D. Koller, C. Gearhart, and N. Kanodia. **Generalizing plans to new environments in relational MDPs.** In International Joint Conference on Artificial Intelligence (IJCAI-03), 2003.

//**Feb 2008**// Ilachinski, A. **Irreducible Semi-Autonomous Adaptive Combat (ISAAC): An Artificial-Life Approach to Land Combat.** Research Memorandum CRM 97–61. (1997) Alexandria: Center for Naval Analyses.

K. Koperski, J. Adhikary, and J. Han. **Spatial data mining: Progress and challenges survey paper**. In Proc. ACM SIGMOD Workshop on Research Issues on Data Mining and Knowledge Discovery, Montreal, Canada, 1996

Kenneth D. Forbus, James V. Mahoney and Kevin Dill, **How Qualitative Spatial Reasoning Can Improve Strategy Game AIs**, Interactive Entertainment 2002

Ana Paula Dutra De Aguiar el al., **Modelling Spatial Relation by Generalized Proximity Matrices**

Sriatsan Laxman and P S Sastry, **A Survey of Temporal Data Mining**, Sadhana Vol.31, Part 2, April 2006, pp. 173-198

Frederik Schadd, **Hierarchical Opponent Models for Real-Time Strategy Games**, 2007

Darse Billings et al., **[|Game-Tree Search with Adaptation in Stochastic Imperfect-Information Games]**,[| Lecture Notes in Computer Science] Volume 3846/2006, p21-34

C. Fairclough, M. Fagan, B. Mac Namee & P. Cunningham, **Research** **Directions for** **AI in** **Computer** **Games**, Proc. of the 12 Irish Conference on AI and Cognitive Science, 2001.

Some of my annotations of these papers are in this page.

Suggested Reading
[|Apply AI 2007 Roundtable Report] --- from [|AiGameDev.com]