The Metaphorical Approaches To Knowledge Representation

Knowledge representation is a prerequisite for the artificial intelligence. In fact, both artificial intelligence and epistemology have one of the most fundamental inquiries:how is the knowledge possible? Since Kant's "Copernican revolution" accomplished "the epistemological turn" of the western philosophy, this inquiry had become a main topic in philosophical study. Until the twentieth century,"the linguistic turn" started to focus on the symbolic systems and ideographic systems. Metaphors widely existing in scientific discourse and daily language as a way of knowledge representation were gradually becoming one of the main subjects of the philosophical language. In order to make all knowledge representation projects fully express the sense of the world, a related semantic theory is surely required, which characterizes the corresponding relationships between the symbols and its represented things. Moreover, Semantic rules not only mean a kind of conversion from the knowledge to "semantic" representation, but the alleged object of the representation must also be specified, which requires a standardized operating symbolic system.This paper, based on the symbolic system and its internal structure mechanism of the knowledge representation, will review the metaphorical characterization in the knowledge representation, so that we can further discuss the significance of the metaphor as a way of the knowledge representation. To start with the semantic system of the knowledge representation, the first chapter aims at explaining the static linear models of metaphorical representation by exploring John· Sowa's Conceptual Graph Theory. However, Conceptual Graph seems still not clearly to specify the relationships between concepts. Therefore, the second chapter will emphasize on the nature and the structure of the semantic hierarchy, because the semantic hierarchies is in nature the links and the relationships between the concept nodes, which can vividly explain the hierarchical structure between concepts in metaphorical representation-triangle hierarchy. So far, we have discussed the structure of the semantic hierarchy itself in detail and specified the static model of the metaphorical representation mechanisms. The following third chapter, with the knowledge representation in the artificial intelligence as its foundation, will introduce the "Interaction Theory" on metaphors to clarify the relationships among language foundation, type hierarchy and conceptual graphs, so that we can further discuss the dynamic structures of semantic hierarchy and obtain an in-depth understanding on the language changes because of semantic hierarchy changes in different situations. Thus, we know that these changes just show the language's very nature of open texture, which also provides us with a clue in grasping the programming modes of this open texture.