论文部分内容阅读
以中国象棋为研究对象讨论象棋知识的表示问题.针对中国象棋的特点,在文中采用组合来描述进攻,并根据组合的不同特点,提出了可用于表示中国象棋知识的物理相关模式和逻辑相关模式的概念,以及由组合得到模式的方法.为在对弈过程中能得到合适的棋步,还提出了目标驱动和数据驱动两种模式的匹配方法,并认为有经验的棋手是以数据驱动为主的.为了提高模式匹配的效率,文中还将模式条件区分为主要条件(即模式特征)和次要条件,并提出了模式匹配步骤和棋步的复杂性概念.本文所提出的知识表示方法,已成功地用于中国象棋的下棋程序.
This paper discusses the representation of chess knowledge by taking Chinese chess as the research object.According to the characteristics of Chinese chess, this paper uses the combination to describe the attack, and according to the different characteristics of the combination, presents the physical and logic related patterns that can be used to represent Chinese chess knowledge As well as the way to get the pattern by the combination.For the purpose of getting the correct move in the game, we also put forward the target-driven and data-driven matching methods and think that the experienced players are data-driven In order to improve the efficiency of pattern matching, the article also divides the pattern conditions into the main conditions (ie, pattern features) and the secondary conditions, and puts forward the concept of complexity of the pattern matching step and the move step.The proposed knowledge representation method , Has been successfully used for Chinese chess chess program.