基于HNC的命题逻辑推理方法

• 作 者: 李颖 田永昌
• 机 构: 装甲兵工程学院科研部,装甲兵工程学院研究生管理大队,
• 出 处: 装甲兵工程学院学报
• 发表日期: 2013年02期
• 关键词: 自然语言理解 句类分析 命题 推理

针对面向自然语言推理中计算机难以准确提取命题并进行符号化的难题,运用HNC(Hieratical Network of Concepts)理论中的语言逻辑概念与句类分析技术,从概念的角度出发,对自然语言描述的推理问题进行理解处理,使计算机能自动提取出推理条件中蕴涵的命题,然后将命题进行公式化处理,最后得到用自然语言描述的结论命题。试验结果表明:该方法能准确理解自然语言中的语义逻辑关系,完成命题逻辑推理。

Aiming at the difficult problem that the computer can not accurately extract and symbolize in the natural language reasoning, this paper applies the language logic concept and the sentence class analysis technology in the theory of HNC (Hieratical Network of Concepts) to conceptualize the natural language from the perspective of concept The problem can be understood and processed so that the computer can automatically extract the propositions contained in the inference conditions, and then formulate the propositions, and finally obtain the conclusion propositions described by the natural language. The experimental results show that this method can accurately understand the semantic logic relations in natural language and complete propositional logic reasoning.