关键词: 定题搜索, 主题爬行, 搜索引擎, 爬行算法, 相关度分析

Abstract: With the explosive growth of the World-Wide Web，to general-purpose crawlers and search engines which pose great challenges.All sorts of special topic search engines are designed for special people and special domains.The web topic information search system（web spider） is the most important part of topic search engine，it collects web pages of special topic and provides users with the result or stores it in index database.Information resource of web is so extensive，how to collect interest content comprehensively and effectively，it is important to web spider research.In this paper，a new crawling strategy block-based topic crawling has been proposed，the experiments show that compared with some traditional algorithms，this algorithm has better performance.It is effective and has high precision.

Key words: topic-specific search, topic crawling, search engine, crawling algorithm, correlation analysis