基于改进协同训练的本体映射方法  

Improved co-training based ontology matching method

作　　者：孙煜飞[1,2] 马良荔[1] 吕闽晖[3] 覃基伟[1] 

机构地区：[1]海军工程大学电子工程学院 [2]中国人民解放军91635部队 [3]海军工程大学装备经济管理系

出　　处：《系统工程与电子技术》2017年第2期459-464,共6页Systems Engineering and Electronics

基　　金：国家部委基金资助课题

摘　　要：针对基于机器学习的本体映射方法存在的人工标记代价高和样本不平衡问题,将本体映射建模为二视图上的协同训练问题,分别从本体模式层和数据层提取特征集。通过对本体概念对进行预匹配,缩小样本不平衡度。分析了传统协同训练方法的局限性,结合主动学习思想,设计了一种考虑样本价值的改进协同训练算法,在每轮迭代中选择更有价值的未标记样本更新训练集。实验结果表明,该方法学习效率更高,同时能取得较好的本体映射结果。Aiming at the problem of high labeling cost and imbalanced data exists in the machine learning based ontology matching, the model ontology matching is taken as an co-training problem on two views, which respectively extracts feature sets from the ontology schema level and the data level. Through pre-matching ontologies＇ concept pairs, the imbalanced degree of samples is reduced. After analyzing the limitations of the traditional co-training method, in combination with the active learning theory, an improved sample value consid- ered co-training method is proposed, it can choose more valuable unlabeled samples to update training sets in each iteration. Experimental results show that the proposed method is more efficient and achieves better com- petitive ontology matching results.

关 键 词：本体映射 协同训练 主动学习 不平衡数据 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]