面向对象的命题逻辑和分布式推理

Object Oriented Proposition Logic and Distributed Deduction

命题逻辑是人工智能和知识工程的基础之一,然而传统命题逻辑却没有体现面向对象的思想,并且难于适应目前的分布式计算环境。本文用对象封装命题变元,实现了传统命题逻辑和面向对象思想的结合;本文引入命题关联类/对象来封装产生式,不但显式地表证了事物之间的命题逻辑关系,还使系统形成一种网络拓扑结构,实现了知识的分布式存储;本文还提出了真值传递和规则触发机制,通过它们可以将推理分布在网络的各个节点完成,并且实现了知识的自索引:相关的知识都存于本地,因此搜索只发生在一个节点内部,避免了全局搜索。

Proposition Logic is one of the foundations of Artificial Intelligence and Knowledge Engineering. However, traditional Proposition Logic is not designed in conformity to OO ideology. and has become inadaptable to modern distributed computing environment. The paper integrates traditional PL and OO methodology by encapsulating statement variables in objects so that OO technology is applicable to traditional Proposition Logic; and the paper introduces association classes/objects so as to encapsulate statement formulas,by which,the relations among entities are explicitly depicted and the distributed storage of knowledge is hence capable because of the network topology formed- Besides , the paper brings up a true value diffusion process and rule triggering mechanism to enable the implementation of deduction among distributed nodes in the net work, thus .enables a knowledge self-indexing: since relevant knowledge has been stored locally .global search is unnecessary but a local one in stead-