本体稀疏向量凸优化学习算法  

Convex Optimization Algorithm for Ontology Sparse Vector Learning

作　　者：朱林立[1,2] 潘瑜[1] 高炜[3] 

机构地区：[1]江苏理工学院计算机工程学院,江苏常州213001 [2]江苏省机电产品循环利用技术重点建设实验室,江苏常熟215500 [3]云南师范大学信息学院,云南昆明650500

出　　处：《云南师范大学学报（自然科学版）》2016年第5期38-43,共6页Journal of Yunnan Normal University:Natural Sciences Edition

基　　金：国家自然科学基金资助项目(11401519);江苏省机电产品循环利用技术重点建设实验室开放基金资助项目(RRME-KF1612);江苏理工学院自然科学研究基金资助项目(KYY14013)

摘　　要：随着大数据时代的到来,本体模型所要处理的数据量越来越庞大.在这一背景下,本体稀疏学习算法越来越受到重视.文中利用稀疏向量学习得到本体函数,用凸优化模型得到本体稀疏向量.使用平衡函数构建平衡项,给出凸集的选取方法,并通过迭代策略得到平衡函数的表达式.最后,两个仿真验证实验表明算法是有效的.With the coming of the era of big data,the amount of data to be processed in the on- tology model increasingly large. For this reason, ontology sparse learning algorithms are raised more attention. In this paper,the ontology function is obtained in terms of sparse vector learning, and the optimal ontology sparse vector is learned by convex optimization model. The balance term is constructed in view of balance function, the selection methods for convex set are presented,and the balance function is then yielded by virtue of iterative computation. At last,the data from two experiments show that the new algorithm has higher efficiency.

关 键 词：本体 相似度计算 本体映射 稀疏向量 

分 类 号：TP393.092[自动化与计算机技术—计算机应用技术]