八邻域网格聚类的多样性XML文档近似查询算法  

Approximate Query Algorithm Based on Eight-Neighbor Grid Clustering for Heterogeneous XML Documents

作　　者：衡星辰[1] 罗俊颉[2] 郭俊文[1] 覃征[1] 邵利平[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049 [2]陕西省人工影响天气办公室

出　　处：《西安交通大学学报》2007年第8期907-911,共5页Journal of Xi'an Jiaotong University

基　　金：国家重点基础研究发展规划资助项目(2004CB719401);国家自然科学基金资助项目(60542004)

摘　　要：提出了一种基于八邻域网格聚类的多样性XML近似查询算法.首先给出了支持XML文档间语义距离计算的3种编辑操作代价模型,再利用XML文档间的语义距离建立XML文档的向量模型并设计基于八邻域网格的XML文档聚类算法,进而利用聚类过程中得到的物理和逻辑聚类中心对静态有序选择算法的查询评估策略进行优化,这样做只需定位聚类中心所在组群的局部范围,并在该范围内进行目标查询,而无需遍历整个XML数据库,从而快速返回满足用户需求的查询结果.经汽车外形智能化设计实验表明,所提算法的查询速度比静态有序选择算法平均提高了3~4倍.An approximate query algorithm based on eight-neighbor grid clustering for heterogeneous XML documents is proposed. Firstly, the cost models of three editing operations used to compute semantic distance between XML documents are given. Secondly, by using the semantic distance, vector models of XML documents are built, and eight-neighbor grid based clustering algorithm for XML documents is designed. Thirdly, the query evaluation strategy for the static order-selective algorithm is optimized with the physical and logical clustering centers obtained from the process of clustering, thus the cluster center can only be located and queried within a local area of document groups without extending all over XML documents, and the query results satisfying the users＇ requirements can be returned rapidly. The experiments of intelligent design of automobile shape show that compared to static selectivity order algorithm, the efficiency of the algorithm is increased averagely by 3 - 4 times.

关 键 词：多样性 近似查询 语义距离 八邻域 静态有序选择 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]