SBC { Proceedings of SBGames 2020 | ISSN: 2179-2259 Computing Track { Full Papers Investigating Case Learning Techniques for Agents to Play the Card Game of Truco Ruan C. B. Moral Gustavo B. Paulus Undergraduate Program in Computer Engineering Graduate Program in Computer Science Federal University of Santa Maria – UFSM Federal University of Santa Maria – UFSM Santa Maria – RS, Brazil Santa Maria – RS, Brazil
[email protected] [email protected] Joaquim V. C. Assunção Luis A. L. Silva Applied Computing Department Graduate Program in Computer Science Federal University of Santa Maria – UFSM Federal University of Santa Maria – UFSM Santa Maria – RS, Brazil Santa Maria – RS, Brazil
[email protected] [email protected] Abstract - Truco is a popular game in many regions of South situation (a current case problem to be solved). Once a America; however, unlike worldwide games, Truco still concrete problem-solving experience is retrieved from requires a competitive Artificial Intelligence. Due to the memory, materialized as a case base in CBR, the agent limited availability of Truco data and the stochastic and reuses the decision made in the past to solve the current imperfect information characteristics of the game, creating problem. For a case-based agent to perform actions in a competitive models for a card game like Truco is a competitive game, that agent can initially use challenging task. To approach this problem, this work demonstrations of game actions performed by human investigates the generation of concrete Truco problem- players, where demonstrations of various kinds can be solving experiences through alternative techniques of recorded as cases in the case base [5].