Internet上日新月异的大量信息资源充斥 ,这对搜索引擎提出更高要求 .为提高搜索引擎中Spider的智能化 ,引入一种新的数据模型和文档自动检索模型 ,其中文档自动检索主要由关键字提取和抽取文摘两部分组成 .在此模型基础之上构造全新的数据检索方法 ,用模型匹配的方法搜索网页为此模型提供带权值的信息资源 ,为精确描述WEB信息提供了依据 .这种检索方法同时增强了Spider的适应性 ,简化了编程维护工作量 . Due to the large amount of information resources on the Internet, the search engines make more demands.In order to improve the intelligence of Spider in search engines, a new data model and automatic document retrieval model are introduced, in which the automatic retrieval of documents is mainly carried out by keywords And extracted abstracts.This paper constructs a new data retrieval method based on this model and searches the web pages by using the model matching method to provide a weighted information resource for this model and provides a basis for the accurate description of the WEB information. The approach also enhances Spider’s adaptability and simplifies programming and maintenance efforts.