搜索结果优化中基于层次型分类的聚类方法研究

搜索引擎根据特定关键字查询返回的结果,可以基于语义进行分类组织,提高用户查询效率。但分类方法是基于预定义类别的,由于类别不全或更新不及,对于互联网上的信息可能会造成遗漏。本文提出了一种将分类与聚类方法相结合的方法来优化搜索结果,即分类之后,用聚类的方法来处理未被归入任何类别的信息。研究表明,该方法可以兼顾效率和信息完整性。

Clustering Based on Hierarchical Structure Classification for Web Search Results

The results returned by search engine can be categorized into hierarchy structure through ontological approach to improve the efficiency of Web Search. However, such classification is based on predefined categories, which can not be up-to-date at real time. Important information may be omitted in such a mechanism. In this paper, we've designed a system in which we combine the classification and clustering method to optimize the search results, that is, we propose to use clustering algorithm just after the classification process to collect the remaining results. This approach has a better balance between efficiency and information integrity.