基于情境约束的知识个性化推送  

Knowledge Personalized Push Based on Context Constraint

作　　者：王有远[1] 赵璐[2] 张乐恩 WANG Youyuan ZHAO Lu ZHANG Le＇en(Institute of Industrial Engineering, Nanchang Hangkong University, Nanchang, 330063 College of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics, Nanjing, 211106 School of Aeronautical Manufacturing Engineering,Nanchang H angkong University, Nanchang,330063)

机构地区：[1]南昌航空大学工业工程研究所,南昌330063 [2]南京航空航天大学计算机科学与技术学院,南京211106 [3]南昌航空大学航空制造工程学院,南昌330063

出　　处：《中国机械工程》2017年第15期1812-1819,共8页China Mechanical Engineering

基　　金：国家科技支撑计划资助项目(2013BAF02B01);江西省火炬计划资助项目(20151BBE51064);江西省科技支撑计划资助项目(20141BBE53005)

摘　　要：对面向产品设计的情境信息进行语义描述和分析,将情境概念之间的关联信息作为限制性条件,提出了基于情境约束的知识聚类方法。采用贝叶斯理论对聚类知识进行语义标注,对标注概念与需求概念的综合语义相似度和相关度进行计算,扩展了检索概念范围。实验结果表明,该研究方法可有效提高知识查询与推送的精确度和效率。Semantic description and analysis of context informations as restrictive condition between contextual concepts for product designs were presented,and a knowledge clustering method was proposed based on context constraints was proposed.Meanwhile,Bayesian theory was used to semantic annotation of clustering knowledges,and the semantic similarity and relevance of annotation concept and demand concept were calculated to expand the scopes of search.Experimental results show that the method may effectively improve accuracy and efficiency of knowledge query and push in the processes of product designs.

关 键 词：情境约束 约束聚类 贝叶斯理论 语义标注 个性化推送 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]