文本检索综述

A Survey of Text Retrieval

文本检索是最早也是最重要的信息检索形式。本文从基于文字、基于结构、基于用户信息几个方面总结了信息检索中相关度计算的方法。对基于文字的信息检索，本文分别介绍了传统的布尔模型、向量空间模型、概率模型和近年以来兴起的统计语言IR模型。文本检索和其他学科逐渐融合构成当今文本检索的发展趋势，本文主要介绍自然语言处理、数据挖掘技术和文本检索的融合，并介绍了数字图书馆中的一些新的文本检索应用。

Text retrieval is one of the earliest and most important retrieval applications,This paper divides the relevance similarity computation appraches into three parts：text based,structure based and user information based approaches,For text based methods,classic Boolean model,Vector Space Model,Probabilistic Model and a recently proposed model-Statistical Language Modeling IR model are discussed,Many technologies from other research areas have been applied for text retrieval;these areas include natural language processing ,data mining,etc.Some technologies from these areas and some new retrieval applications in Digital ibrary are also discussed.