Research Of Automatic Indexing In Economic Bibliographical Database

This paper introduces the advance of the computer-assisted indexing and word- segmenting technique. it points out the questions about computer-assisted subject indexing and classifying of Chinese document. According to the indexing condition of The index to the National Periodicals and Newspapers in Shanghai Library, the author designed and realized a computer-assisted indexing system of Chinese bibliographical database. This paper explores and discusses the method of how to establish the vocabulary, include stop-lists, keywords dictionary, keywords-thesaurus conversion dictionary, place-name lists, and thesaurus-classification term concordance based on indexing experience. Directed by the theory of integration of library classification and thesaurus, and the theory of retrieval language's compatibility and exchange, the author designed an indexing system that combine subject indexing with classifying. Firstly, the system improved word-segmenting algorithm of the maximum matching method and word by word searching method to make the word segmenting accurate and not redundancy. . Secondly, the system combined word-segmenting indexing with word-assigned indexing to get keywords and subject terms. It also illustrates the effect of different part of text. Thirdly, the system classified mainly by title keywords and abstract. Compare with manual indexing, this system has an advantage over the indexing speed and indexing consistency. The quality of automatic indexing and automatic classifying is close to the manual indexing.