摘要
为准确推断可扩展标记语言(XML)关键字查询中的用户搜索意图,提出了一种新的推断XML关键字查询目标节点类型方法 IXKSNT。在获取目标节点类型时,考虑不同节点类型下关键字出现的频率以及节点类型的层次信息,赋予不同的查询相关度,同时结合目标节点类型应存在实例节点包含所有查询关键字这一因素,推断目标节点类型,以保证查询结果节点含有全部关键字。实验证明,该方法可以更准确地得到目标节点类型,提高查准率。
This paper proposes a novel method to infer target node type in XML keyword search called IXKSNT, which can accurately infer users＇ intention. When IXKSNT gets the different node types, gives them different relevance value based on the frequency of keyword and the level information of node type. Meanwhile, depending on the idea that target node type must have at least one instance node that contain all keywords, which ensures that result nodes will include all search keywords. The experimental evaluation shows that the method can find the target node type more accurately and enhance the accuracy.
出处
《信息技术》
2013年第12期34-38,共5页
Information Technology
基金
国家自然科学基金项目(61170200)
江苏省科技支撑计划(工业)项目(BE2012179)
江苏省普通高校研究生科研创新计划项目(CXZZ12_0229)
关键词
XML
关键字查询
目标节点类型
词频
层次信息
XML
keyword search
target node type
term frequency
location information
