区域划分并行扫描下网络海量元数据查询算法设计  

Design of Network Massive Metadata Query Algorithm under Regional Partition Parallel Scanning

作　　者：黄雄平 罗伟 HUANG Xiongping;LUO Wei(Department of Academic Affairs,Guangdong Polytechnic of Science and Technology,Zhuhai 519090,China;Library,Zhuhai College of Science and Technology,Zhuhai 51904l,China)

机构地区：[1]广东科学技术职业学院,教务部,广东珠海519090 [2]珠海科技学院,图书馆,广东珠海519041

出　　处：《微型电脑应用》2024年第10期26-28,33,共4页Microcomputer Applications

基　　金：2021年广东省高等职业教育教学质量与教学改革工程项目(GDJG2021156);2021年广东省继续教育质量提升工程项目(JXJYGC2021AY0027)。

摘　　要：网络数据包含大量目标对象,导致元数据描述条目较多且标签密集,增大了查询难度。为此,提出基于区域划分和并行扫描的网络海量元数据查询算法。所提算法通过应用子森林理论对海量元数据进行区域划分,以解决元数据中标签过于密集的问题。根据元数据区域划分结果,设计元数据并行扫描算法,同时扫描多个元数据分区,以提高查询速度和整体效率。开发基于并行扫描算法的互联网海量元数据查询程序,根据查询请求参数获得查询结果,实现互联网海量元数据精准查询。实验数据显示,所提算法应用后的元数据查询响应时间最小值为7.25 ms,元数据查询出错率最小值为4.8%,具有较好的元数据查询性能。Network data contain a large number of target objects,resulting in a large number of metadata description entries and dense labels,which increases the difficulty of querying.Therefore,a network massive metadata query algorithm based on re-gion partitioning and parallel scanning is proposed.This algorithm applies sub forest theory to partition massive metadata into regions to solve the problem of label density in metadata.Based on the results of metadata region partitioning,this paper de-signs a metadata parallel scanning algorithm that simultaneously scans multiple metadata partitions to improve query speed and overall efficiency.The Internet massive metadata query program is developed based on parallel scanning algorithm,and the query results are obtained according to query request parameters,and achieve precise querying of Internet massive metadata.The experimental data show that the proposed algorithm has a minimum metadata query response time of 7.25 ms and a mini-mum metadata query error rate of 4.8%,indicating good metadata query performance.

关 键 词：元数据 互联网资源 查询方法 联合查询 内存数据库 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]