自然语言语篇推理是自然语言信息处理、计算机科学中的知识表示和知识推理的重点与难点内容之一。利用广义量词理论、集合论和模糊逻辑相关知识,可以简洁明了地对包含 al-most all 的广义三段论进行形式化的表示和证明。对此有18个事实可以印证。在广义三段论形式化和有效性研究的基础上,可以探讨由两个及以上的广义三段论嵌套而成的自然语言语篇推理。这些创新性成果和研究方法,将为考察包含其他广义量词的广义三段论的形式化、有效性及其相关语篇推理提供一个简洁合理的数学范式。%Discourse reasoning in natural languages is one of the emphases and difficulties for natural language information processing and for knowledge representation and reasoning in computer science. The generalized syllogisms including almost all can be formalized and proved by means of generalized quantifier theory,set theory and fuzzy logic.This is illustrated by 18 facts.Discourse reasoning in natural languages nested by two or more generalized syllogisms can be discussed on the basis of the formalization and validity of generalized syllogisms.These innovative achievements and methods in this paper provide a simple and reasonable mathematical model for the research regarding the formalization and validity of generalized syllogisms including other generalized quantifiers,and the relevant discourse reasoning.
展开▼