부산대학교 도서관

Digitalizing tabletop games for general game playing (GGP) AI research is a continuously growing field. Tabletop Games Framework (TAG) is a framework developed to simplify the process of implementing tabletop board games to digital form. Sushi Go! is a game that combines simultaneous action selection and complete information. This creates a unique combination of mechanics, which presents a new challenge for GGP agents. By implementing Sushi Go! into TAG, we can test different agent's performance using these mechanics and compare them to their existing performances in the other games of TAG. Results of this testing are presented, which display that the framework is capable of implementing Sushi Go! and that the agents perform with mixed results. Further developing heuristics for the agents should prove to increase their performance when faced with these types of games.