融合三支决策与GRA-Orthopair模糊集的知识匹配方法研究  

Research on knowledge matching method combiningthree-way decision and GRA-Orthopair fuzzy set

作　　者：张建华[1] 刘艺琳 温丹丹[1] 杨岚[2] Zhang Jianhua;Liu Yilin;Wen Dandan;Yang Lan(School of Management Engineering,Zhengzhou University,Zhengzhou 450001,China;School of Mechanical&Power Engineering,Zhengzhou University,Zhengzhou 450001,China)

机构地区：[1]郑州大学管理工程学院,郑州450001 [2]郑州大学机械与动力工程学院,郑州450001

出　　处：《计算机应用研究》2021年第7期1967-1972,共6页Application Research of Computers

基　　金：国家社会科学基金资助项目(19BTQ035)。

摘　　要：为缓解用户“知识迷向”问题,提升知识应用及创新效益,提出了融合三支决策与GRA-Orthopair模糊集的知识匹配方法。首先基于CRITIC-熵权法进行权重配置,并融入博弈论方法增强客观赋权实效;而后,采用FCM聚类算法横向压缩匹配空间;进而,借鉴三支决策思想实施预分类,同时考虑知识间的关联度,构建基于灰色关联度的Orthopair模糊相似度模型,并据此计算用户需求与案例知识间的视图相似度,从而确定匹配结果。实验结果表明,该算法的知识匹配精度及效率优于传统匹配算法,提升了匹配效果。In order to alleviate the user’s“knowledge obsession”problem and improve knowledge application and innovation efficiency,this paper proposed a knowledge matching method that combined three-way decision and GRA-Orthopair fuzzy sets.First,this paper used CRITIC-entropy method for weight configuration,and integrated sgame theory methods to enhance the objective weighting effect;then,it used the FCM clustering algorithm to compress the matching space horizontally;and then,it used the three-way decision ideas to implement pre-classification,considered the inter-knowledge relevance degree,constructed Orthopair fuzzy similarity model based on gray relevance degree,and calculated the view similarity between user needs and case knowledge based on this,so as to determine the matching result.Experimental results show that this algorithm is better than traditional matching algorithms in knowledge matching accuracy and efficiency,and improves the matching effect.

关 键 词：知识匹配 博弈论 FCM 三支决策 灰色关联分析 Orthopair模糊集 

分 类 号：TP393.04[自动化与计算机技术—计算机应用技术]