CAN: Case-based reasoning in an adversarial non-player character

• Main Authors: Africa, Anne Marie M., Cortez, Karmela Angela G., Fontanilla, Gian Kristian A., Go, Paul Michael O., Cheng, Danny C.
• Format: text
• Published: Animo Repository 2022
• Subjects: Artificial intelligence; Games of strategy (Mathematics); Game theory—Computer programs; Computer Sciences
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/5168
• Institution: De La Salle University

Game development, which was primarily driven by the desire to achieve realistic graphics, is now turning to other innovations such as applying Artificial Intelligence (AI) techniques (Cunningham, et al., 2003). AI algorithms of heuristic nature such as the A* path finding algorithm, A-Life/flocking algorithms, and Fuzzy State Machines (FSMs) are well understood in the context of computer games (Cunningham et al, 2001). However, such algorithms only make use of simple decision making and still lack the ability to learn (Bernon). One type of game that requires this kind of learning is real-time strategy games. This research intends to design an adversarial Non Player Characters (NPC) that learns strategies in a real time strategy (RTS) game.