A Study on the Structure of Korean Knowledge Database

A Study on the Structure of Korean Knowledge Database1 Yude Bi, Binhong Wu, Jianguo Xiong PLA University of Foreign Languages, P.O.Box 036-50, 471003 China biyude@gmail.com, 3241122lm@sina.com, jianguoxiong@163.com Abstract. Knowledge is most often and directly expressed with natural language. Therefore, the representation and acquisition of knowledge play a key role in studies on information processing and natural language interpretation. Also, they are the principal issues in establishing a knowledge database. Only highly formalized descriptive systems or knowledge databases can be processed by computers. On the basis of the descriptive mechanisms of lexicalist theory, the present paper tries to provide a detailed description and integration of the syntactic, semantic and other information of Korean, in the form of lexical structures. Keywords: Korean, knowledge database, lexical structure 1 Introduction Studies on knowledge projects or intelligent systems have developed to such an extent that exploration into knowledge acquisition and intelligent simulation has become possible. Now, we must pay closer attention to basic theoretical studies, especially in the field of syntax and semantics. These studies present themselves as pioneering research topics in the field of language information processing and have attracted the attention of many experts both at home and abroad. Focus is placed on the description and acquisition of syntactic and semantic knowledge. The present study is intended to construct a language knowledge presentation system for Korean information processing. The core of this system is a knowledge database encompassing both syntactic and semantic information. Doubtlessly, the design of this database plays a crucial role, because it has a direct impact on the operation and finally the overall quality of the system. 2 Theoretical background As stated above, syntactic and semantic information constitute the core of a knowledge database for information processing, therefore, finding an efficient way for the description of such information determines the final adequacy of the database. This calls for a clarification of the relations between syntax and semantics. On the one hand, semantics puts constraints on the range of available syntactic expressions. And syntax must be loyal to semantic content for the proper communication of the speaker’s ideas and intentions. On the other hand, semantics must follow syntactic rules in order to get the ideas properly expressed. The form of a sentence is subject to both syntactic and semantic constraints. For syntactic constraints between constituents, we have relations such as subject-predicate, verb-object, modifier-head, etc. For semantic ones, we have relations between action and agent, patient, instrument, time, place and others. Such semantic relations are addressed as syntactic-semantic relations. Both syntactic and semantic structures are formed in the derivation or composition of a sentence. In language use, meanings are always expressed with certain syntactic structures or forms. Syntactic meaning is no exception. 11 This work was supported by the National Social Science Foundation of China (No. 05BYY019) and Brain Korea 21 Project, The school of information technology, KAIST in 2005. 95 However, semantic and syntactic structures do not have one-to-one correspondence. One syntactic structure may have several semantic relations; one semantic relation …

Content truncated to 500 words