数据库模式的主动在线匹配方法  

Active online matching approach for database schemas

作　　者：庄莉 陈又咏 黄双双 丁阳 张照 ZHUANG Li;CHEN Youyong;HUANG Shuangshuang;DING Yang;ZHANG Zhao(Fujian Yirong Information Technology Co.,Ltd.,Fuzhou 350003,China;School of Electronic and Information Engineering,Xi’an Jiaotong University,Xi’an 710049,China;Beijing SGITG-Accenture Information Technology Co.,Ltd.,Beijing 102211,China;Beijing China-Power Information Technology Co.,Ltd.,Beijing 102211,China)

机构地区：[1]福建亿榕信息技术有限公司,福建福州350003 [2]西安交通大学电子与信息学部,陕西西安710049 [3]北京国网信通埃森哲信息技术有限公司,北京102211 [4]北京中电普华信息技术有限公司,北京102211

出　　处：《现代电子技术》2022年第1期34-39,共6页Modern Electronics Technique

基　　金：国网信通产业集团两级协同研发企业级办公软件自主可控关键技术研究和应用示范项目。

摘　　要：现有研究中常使用批量学习实现数据库间的模式自动匹配,但实际应用中模式匹配任务具有经验积累的增量性特征,难以为批量学习一次性标定大量学习样本。为此提出了一种数据库模式的增量式动态匹配方法,针对真实应用场景下模式匹配任务序贯到达经验动态增加的特点,采用集成学习与主动在线贝叶斯匹配算法增量学习数据库模式间的关联关系,在模式匹配样本渐进积累过程中,实现数据库模式的自动匹配。实验结果表明,给定相同初始训练样本,相比批量学习方法,所提方法能够将模式匹配F_(1)-score提高5%~33%。In the existing researches,batch learning is generally used to achieve automatic schema matching of databases.However,in actual application scenarios,schema matching task has the incremental feature in the accumulated experience,so it is difficult to calibrate a large number of learning samples for batch learning at one time. To this end,an incremental dynamic matching method for database schemas is proposed. In view of the sequential arrival characteristics of scheme matching tasks and dynamic increase of experiences in real application scenarios,the integrated learning and active online Bayesian matching algorithm are adopted. The relationship between incremental learning database schemas can realize automatic matching of database schemas during the gradual accumulation of schema matching samples. Experiments show that,given the same initial training samples,the proposed method can improve the scheme matching F_(1)-score by 5%~33% in comparison with the batch learning algorithm.

关 键 词：数据库 模式匹配 朴素贝叶斯模型 增量学习 随机森林 批量学习 

分 类 号：TN911.1-34[电子电信—通信与信息系统]