Google Alerts brought this extensive masters paper to my attention, by Christopher Alan Ballinger. The paper explains what expert systems are and has a lot of DRL examples to follow.
I’d be interested to see this code published online for others to play with.
“In this project, we describe the design decisions and principles behind the arti cial intelligence (AI) for a multiplayer online role playing game, and our use of an expert system to implement it. We explain how we organize AI rules into les, how those rules are assembled from a database, how AI is assigned to entities, the di erent types of AI, the di erent phases of AI, and how we manage facts used by AI. We also review some of the history behind the Mythic project, where it is headed, what an expert system is and why we chose to use one for our project. The result of our project is a design that allows us to have diverse AI behavior and exibility to reuse code to create new behaviors, but may prove to be ineffcient if implemented on systems with many players or many instances of AI running.”