Abstract: To improve the correct-rate and completeness-rate of search, a topic-specific meta-search engine is designed. A bogus crawler is invented, which collects information by the normal search engines, so that the search-area is wider than the normal search engine. The feedback mechanism is adopted and the search-history of user is considered, which make the search result is more imminent to the purpose of the user. Owing to the strategy of topic-specific and mending the arithmetic of similitude-degree of the texts, the correct-rate is improved. Both the correct-rate and completeness-rate of searching are improved, the response time is decreased as well, at the same time, the request of capability of the server is reduced.

Key words: meta-search, topic-specific, search engine, bogus crawler

摘要： 为提高搜索的查准率和查全率，设计一个主题式的元搜索引擎和一个类似于爬行器的伪爬行器，通过调用通用搜索引擎采集信息，查全率高于通用搜索引擎。利用反馈机制，参考用户查询历史记录，搜索结果更加接近用户的要求。通过采用主题式策略，改进文档相似度算法，提高分类的正确率和搜索引擎的查准率与搜索范围，同时减少系统响应时间，降低对服务器性能的要求。

关键词: 元搜索, 主题式, 搜索引擎, 伪爬行器

CLC Number: 

• TP311