Publication: Automatización de la negociación en un juego de mesa con aproximación a boardgame.io
Loading...
Identifiers
Publication date
2021
Defense date
2021-07
Authors
Advisors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Una manera de utilizar la IA para simular problemas del mundo real es aplicarla a los juegos de mesa, ya que estos permiten simplificar dicho mundo por medio de unas reglas y/o casillas determinadas, de forma que se puedan abordar en un entorno controlado. Se ha investigado bastante en este campo, pero la mayoría de algoritmos se han desarrollado para juegos de suma cero, en los cuales la suma de las ganancias de unos jugadores es igual a la suma de las pérdidas de los demás, a pesar de que la mayoría de problemas del mundo real son de suma distinta de cero, ya que el beneficio de una parte no suele ser igual a la perdida de la otra, y/o por permitir algún tipo de negociación.
En este proyecto se desarrolla un juego de mesa de suma distinta de cero que permite negociar, y se automatiza dicha negociación en base a las posibles situaciones que se pueden dar en el juego, programando para ello un Sistema Experto Basado en Reglas con CLIPS.
Además, se realiza una aproximación teórica a boardgame.io de forma que se indican los pasos que se deben seguir si se quiere implementar el juego desarrollado en la herramienta, además de lo que hay que tener en cuenta si se quiere añadir la solución de la automatización de la negociación al juego una vez éste se haya implementado.
A way to use AI to simulate real world problems is to apply it to board games due to the fact that they allow to simplify this world by some determined rules, in a way that can be approached in a controlled environment. Research in this field has been made but most of the algorithms have been developed for zero-sum games, in which the sum of gains of some players is equal to the sum of losses of the rest of players, despite most of real world problems are non zero-sum, because the benefit of one part is not usually the same as the losses of the other and/or for allowing some type of negotiation. In this project a non-zero sum game with negotiations is developed and this negotiation is automated in base of the possible situations that can be possible in the game, programming for it a CLIPS Rule Based Expert System. Also, a theoretical approach to boardgame.io is made in a way that steps to follow for implementing the game developed in the tool is given, apart from what has to be taken into account if implementing the solution of automation of negotiations is intended once the game is implemented.
A way to use AI to simulate real world problems is to apply it to board games due to the fact that they allow to simplify this world by some determined rules, in a way that can be approached in a controlled environment. Research in this field has been made but most of the algorithms have been developed for zero-sum games, in which the sum of gains of some players is equal to the sum of losses of the rest of players, despite most of real world problems are non zero-sum, because the benefit of one part is not usually the same as the losses of the other and/or for allowing some type of negotiation. In this project a non-zero sum game with negotiations is developed and this negotiation is automated in base of the possible situations that can be possible in the game, programming for it a CLIPS Rule Based Expert System. Also, a theoretical approach to boardgame.io is made in a way that steps to follow for implementing the game developed in the tool is given, apart from what has to be taken into account if implementing the solution of automation of negotiations is intended once the game is implemented.
Description
Keywords
Inteligencia artificial, Sistema experto basado en reglas