摘要
针对当前海量信息存储对等网络系统中资源搜索技术效率较低的问题,提出了一种采用k-均值聚类分析的高效搜索模型.该模型利用资源描述框架(RDF)描述的元数据进行聚类分析,使得资源的搜索由全局变为局部,从而有效地提高了资源搜索效率;采用动态优化排序技术显著提高了查询的速度.通过子网分裂算法和节点备用算法增强了模型的可扩展性、安全性和可靠性.仿真结果表明,所提模型在查找时延和平均路径方面均比传统搜索模型更加高效、便捷.
An efficient search model using k-means clustering analysis is proposed to improve the low efficiency of resource retrieval technology in peer-to-peer net with mass information.The metadata described by RDF framework is used to perform cluster analysis of resources and the search range of resources is narrowed from global to local so that the model can enhance the efficiency of resource retrieval effectively,a dynamic optimization technique is adopted to significantly improve the inquiry speed.Moreover,the use of the subnet division algorithm and the node backup algorithm enhances the scalability,safety and reliability of the model.Simulation results and comparisons with traditional retrieval models show that the proposed model is convenient and has higher resources searching efficiency in search delay and average path.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2012年第10期55-59,共5页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(61073196)
陕西省自然科学基础研究基金资助项目(2011JM8026)
关键词
海量信息存储
聚类分析
元数据
搜索技术
mass information storage
clustering analysis
meta data
retrieval technology
