摘要： 根据汉字在文献中使用的频度和分布规律，利用图书分类的原则，提出了基于汉字在成组（连续索引号）文献中出现的集合索引方案。与基于单值表达的常规索引方案作了比较，并对一类短语料进行了实验，给出了部分比较结果以及索引调整的一些分析结果。模拟图书分类中汉字非均匀分布的特点，从理论上分析文献索引各参量间的关系。分析表明，集合索引法与单值表达索引相比，能够显著提高素引效率，减少索引空间，还可进行随机动态调整。

关键词: 集合索引, 单汉字索引, 汉字连续度

Abstract: Chinese word has its own usage frequency rule in the text, slightly different from English word.According to the classification of text, this paper proposed the set-based approach on Chinese word index.Following the results of some experimenis on comparing with the single-value method, the set-based method can reduce index space and accelerate retrieve speed, its efficiency can be enhanced later.In this paper, an analytical model was established for study purpose.

中图分类号: 

• TP391.3