论文部分内容阅读
为了使Prolog推理引擎可以从一定程度的语义上来区分子句,通过概念图的形式对Prolog程序的子句建立本体.子句本体的概念模型通过其谓词参数和头子目标来建立,而头子目标是指出现在规则体首部并且位于任何谓词调用之前的子目标集合.所提出的方法把一个Prolog程序转换成包含其子句本体的Prolog+CG程序.实验表明,通过对Prolog子句建立本体,可以尽早地去除与当前目标明显不匹配的子句,从而使得目标求解树的规模减小.在概念图的描述形式下,子句本体使得Prolog程序的语义更清晰可读,在某些情况下能明显加快程序的求解过程.
In order for the Prolog inference engine to semantically distinguish clauses to a certain degree, the ontology is built on the clauses of the Prolog program in the form of a concept map. The conceptual model of the clause ontology is established by its predicate parameters and the head target Refers to the set of subobjects that are now at the head of the rule and precede any predicate call. The proposed method transforms a Prolog program into a Prolog + CG program containing its clause ontology. Experiments show that by establishing ontologies for Prolog clauses, To remove the clauses obviously inconsistent with the current target, so that the size of the target solving tree is reduced.In the description of the concept map, clause ontology makes the semantics of Prolog program more readable and in some cases obvious Accelerate the process of solving the process.