Abstract: To improve the accuracy of Web service and the query experience of users,document processing technique and query processing technique are hot research topic in Web service search.For this reason,an approach which integrated document processing technique and query processing technique for Web service search was proposed.By considering the relationships between documents and the normalized framework,the re-rank documents in searching list based on the initial results was realized.The candidate terms from expansion terms set based on top documents were selected to expand the query.The experiments were conducted on NTCIR dataset,and the results showed that the proposed method could better improve precision for top results.

Key words: Web service search, document processing technique, query processing technique, normalized framework, expansion terms set

摘要： 为了提高Web服务搜索结果的准确性和改善用户的查询体验,使用文档处理技术与查询处理技术一直是Web服务搜索研究中的热点课题。为此,本文提出一种结合两种技术的Web服务搜索方法。考虑文档与文档之间的关系,通过正则化框架,融合第一轮检索结果分数,实现搜索结果列表中文档重排序;基于排名靠前文档构建扩展词集合,从构建扩展词集合中选取候选扩展词进行查询词扩展。在NTCIR数据集上的实验结果表明,该方法能够有效提高Web服务搜索中排名顶端结果的准确率。

关键词: Web服务搜索, 文档处理技术, 查询词处理技术, 正则化框架, 扩展词集合