利用分类法和主题法改善搜索引擎的性能

Improve the Performance of Search Engines by Using Classification and Subject Indexing

本文对 Ｉｎｔｅｒｎｅｔ上的检索工具搜索引擎的工作机理和性能进行了全方位的探讨，指出了常见搜索引擎信息检索缺点。同时还指出了Ｍｅｔａ搜索引擎、智能搜索引擎和代理搜索引擎虽然提高了网络信息检索的质量，但由于分类体系不统一，类目划分标准模糊，因而仍然未能从根本上改变搜索引擎主要以关键词（自然语言）作为检索入口的现状，不能实现分类检索与主题检索的一体化。要实现搜索引擎信息检索的突破，笔者认为有必要运用情报检索语言的理论和方法来完善因特网搜索引擎的性能，实现分类、主题一体化的检索机制，克服分类检索语言单纯以学科聚类、主题语言单纯以事物聚类的局限性。

The main purpose of this note is to discuss the principles and capabilities of the Internet search engines all-round, point out the fault of the search engine's information retrieval. Although Metasearcher, Intelligent-searcher and agent -searcher can improve the information retrieval quality on network, because of the classification system don't unified, so this is a fact that search engines can not change the retrieval interface by using the keyword (natural language), and is hard to satisfy the high quality retrieval. Therefore, the theory and methods of the information retrieval language are provided to improve the performance of search engines, achieved the intergration of the classification and subject.indexing.