The Research And Practice Of The Conceptual Graphs Knowledge Representation

Paper studied the traditional methods of knowledge representation,such as predicate logic,semantic networks,production,framework,state-space and object-oriented conceptual graphs knowledge representation.Paper analyzed and compared the advantages and disadvantages of these methods,then gived out the characteristics of a comprehensive knowledge representation method.Paper researched the important concept lattice theory of conceptual graphs and analyzed the expressed of conceptual graphs in the form of predicate.On the basis of study of the relationship between conceptual graphs and matching rules,paper gived details of two commonly used reasoning algorithms based on the conceptual graphs, such as projection matching and maximal connection mating.The paper extracted designing regulation combining with the conceptual graphs,and then realized the solution of composite function and inverse function by Visual Prolog. Paper described the concept of bijective,injective,surjective function by the conceptual graphs,then given out the solution and the some examples of their applications.The paper studied the application of conceptual graphs in semantic representation, including the conceptual graphs-based semantic analysis:Chinese word segmentation and semantic chunk,paper presented the simple sentence and complex sentence semantic analysis algorithm base on conceptual graphs and their examples.Paper in-depth studied the application of hash function in the search engine,OR mapping analysis based on the conceptual graphs and isomorphism proof based on conceptual graphs.