一种基于节点语义相关性的XML关键字查询算法

An XML Keyword Query Algorithm Based On Node Semantic Relevance

现有的XML关键字查询算法,通常只考虑节点间的结构信息,以包含关键字匹配节点的子树作为查询的结果,而节点间的语义相关性一直没有被充分利用。这也是导致现有查询算法的结果中普遍含有大量语义无关的冗余信息的主要原因。在该文中,我们首先对查询关键字的环境语义及节点间的语义相关性进行了定义,在此基础上,提出了一种新的关键字查询算法,寻找语义相关单元作为关键字查询的结果。这样获得的查询结果,一方面不含语义无关的冗余信息,另一方面也与用户的查询意图更加匹配。实验表明,该文提出的算法在查询效率和精确性上都有较大改进。

In general,the existing XML keyword query algorithms only consider the structural relationships among XML nodes and return the subtrees including keyword matching nodes as the query results.The semantic relevance is not fully used,which is the main reason that leads to the semantic irrelevance of the query result.In the paper,the context semantics of query keywords and semantic relevance among XML nodes are defined firstly.Based on the definitions,a new keyword query algorithm which finds the semantically relevant units as query results is proposed.Using the algorithm,the query results do not include semantically irrelevant information and really match user's query intention.The accuracy and efficiency of our algorithm are confirmed by extensive experiments.