基于HIPN网的推理算法研究

A Logical Inference Algorithm on HIPN

基于HIPN网的知识表达有知识的长效性和避免重复运算两个优势。提出了一个基于HIPN网模型的逻辑推理算法,该算法结合经典的归结证明策略,并充分利用归约过程中已经得到的中间结论。与现有T-不变量和线性代数的求解算法相比,大幅减少了重复推理过程,有效地提高了推理效率。

Representing knowledge using HIPN can make full use of the two advantages owned by HIPN. longevity of the knowledge and no duplicated running of the rules. We invent an algorithm to realize logical inference on HIPN. The main principle of the new algorithm is to make full use of the intermediate results obtained during reasoning process,so as to reduce the duplicated reasoning steps that may occur in other existed methods. The new methods can reach the goal sentence by fewer steps and so,is more efficient.