文献标引专家系统(DIES)雏型的研究和实现

APPROACH AND IMPLEMENTATION OF A PROTOTYPE SYSTEM FOR THE DOCUMENT INDEXING EXPERT SYSTEM(DIES)

文章探讨了在计算机标引系统中应用专家系统理论和方法的意义和可能性，并初步实现DIES（Document Indexing ExpertSystem）的设计。这一系统着力于揭示主题词之间的各种语义联系，依托于GRA书目数册库所提供的标引数据构成DIES知识库，从而解决了构造专家系统时所面临的知识获取、知识表达等一系列问题，为建立智能型计算机标引系统提供了新的尝试。介绍TDIES系统的结构、功能和程序实现。

This paper approaches the significance and feasibility of the application of expert system theory and methods in the computerbased indexing system. The design of DIES （Document Indexing Expert System） is primarily implemented. The system focuses on revealing various kinds of semantic relationships among the subieet terms. Based on indexing data provided by the NTIS bibliographic database, a DIES knowledge base is established. Problems on knowledge acquisition and know/edge presentation in developing such an expert system are also solved, with brief description of structures, functions and programming of the DIES system, a new trial for establishing an intelligent type of computer-based indexing system is made.