Some problems in controlled thesaurus indexing are pointed out: low indexing efficiency, relation net among the descriptors and new terms aren t developed, and uncertain coordinate indexing. Semiautomatic canonical control method for natural language term after creation of database is raised. The approximate match algorithm among the terms and the building method for U , UF , BT , NT and CT relation are studied.
The article explained the relation between word group and words as the form of chain and ring. The saving of controlled words in ring form not only reduced greatly the memory capacity for the thesaurus in the machine, but also solved the extraction of indexing words by the means of ring. This approach enhanced the processing speed of indexing words, and thus the subject indexing was realized.
A multi-language and multi -function thesaurus management system written by ISIS/Pascal is described. This management system can be used not only to compile thesaurus but also to produce printed text file of thesaurus with all the artistry of a master typesetter automatically.
In this paper, an external sort algorithm of a database file is presrnted. The retrieval efficiency will obviously increase when this algorithm is applied to sort connection of documents.
The design ideas and techniques used in Peking University Library Automation Integrated System (PULAIS) are introduced in this paper, which will be of great reference value to the implementation of library automated systems.
This paper discusses the concepts of communication and communication industry, the development of the world communication industry and the trend of connunication technology as well as the present state of China' s communication industry and its future policy.
