关键词: 可扩展标记语言（XML）, 关键词搜索, 多样化

Abstract: Results of keyword search on extensible markup language (XML) documents are confronted with the problems of high volume, being homogenous in semantics and difficulty in differentiation. To solve these problems, this paper proposes a novel diversification-based method. It first defines the prototype of a search result to express the semantics of the result. Based on the characteristics of result prototype, it defines the interestingness of a prototype and the distance between prototypes. It then diversifies prototypes using these measures. The paper goes further to propose a new method to organize the search results of an XML keyword query, i.e., clustering the search results based on the diversified prototypes. The method can solve the above-mentioned problems. Experimental results verify that the methods are effective.

Key words: extensible markup language (XML), keyword search, diversification