| The 21st century is the era of knowledge economy.With the continuous advancement of global education information reform,artificial intelligence technology(AI)is gradually widely used in the field of education.Knowledge representation is one of the most important problems in AI.At the same time,we must realize that education is the development of "subject person",and students’ learning is the occurrence of knowledge,that is,we should pay attention to the two problems of how knowledge is produced and how knowledge is developed.Guided by Piaget’s genetic epistemology,this study aims to achieve a new form of knowledge representation that separates representation from operation.With the help of linguistics and computational linguistics,this study constructs a theoretical system of category map representation and logically verifies its self consistency.First of all,it combs and analyzes the characteristics and theoretical basis of the previous knowledge model,and finds that it has many limitations,mainly in ignoring the subject of cognition,and confusing the content and operation with the knowledge model represented by "program + content".The reason lies in the lack of the guidance of philosophical epistemology,and blindly seeking the help of psychology,computer science and other downstream disciplines.Secondly,it traces back to Piaget’s genetic epistemology,the latest achievement of philosophical epistemology.With the help of Piaget’s "two categories"-Physical category and logical mathematics category,it corresponds to concept category and logical mathematics category in language.The former provides content style stipulation,while the latter provides rule style stipulation.It puts forward a new knowledge model,namely "knowledge thinking" knowledge Knowledge model is also a new way of knowledge representation,which is the "representation system of category map".In terms of language,the representation system of category map can be divided into semantic category map and grammatical category map.This study expounds the representation system from the theoretical level.In the semantic category map,lexical items,as the expression of concepts,focus on the semantic content and semantic structure represented by lexical items,and with the help of relevant research in Linguistics and other fields,new semantic structure types are divided;in the grammatical category map,complete grammatical structure is taken as the basic node,subject predicate object structure as the root node,and the "tree" grammatical structure is constructed,so as to integrate semantics and grammar Separate representation and realize representation in the sense of prior category.Finally,the author puts forward the assumption of the association between semantic category and grammatical category.On the one hand,from the perspective of representational meaning,grammatical rules need to be explained by lexical meaning,and grammatical rules also affect the evolution of lexical meaning.The author puts forward some solutions to the collocation of words.On the other hand,from the perspective of generative meaning,semantic and grammar are related in pragmatics.In this way,the separation of representation and operation also provides a new guiding idea for the implementation of computation. |