基于区间直觉模糊数的群决策模型及其在社交网络中应用  

A group decision-making model based on interval-valued intuitionistic fuzzy numbers and its application on social network

作　　者：陈晓红[1,2,3,4] 赵翠翠[1] 杨立[2,3,4] CHEN Xiaohong ZHAO Cuicui YANG Li(School of Business, Central South University, Changsha 410083, China Mobile E-business Collaborative Innovation Center of Hunan Province, Hunan University of Commerce, Changsha 410205, China Key Laboratory of ttunan Province for Mobile Business Intelligence, Hunan University of Commerce, Changsha 410205, China Institute of Big Data and Internet Innovation, Hunan University of Commerce, Changsha 410205, China)

机构地区：[1]中南大学商学院,长沙410083 [2]湖南商学院湖南省移动电子商务协同创新中心,长沙410205 [3]湖南商学院移动商务智能湖南省重点实验室,长沙410205 [4]湖南商学院大数据与互联网创新研究院,长沙410205

出　　处：《系统工程理论与实践》2017年第7期1842-1852,共11页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金委重大国际合作项目(71210003);国家自然科学基金重点项目(71431006);湖南省自然科学基金青年科学基金(2016JJ6050);湖南省教育厅优秀青年基金项目(16B145)~~

摘　　要：基于经典熵方法的局限,提出一个新的区间直觉模糊熵计算方法.考虑社交网络数据高度动态及非结构化的特性,引入区间直觉模糊思想,创新性地将社交网络用户影响力量化评价转化为模糊多属性群决策问题,提出基于区间直觉模糊数的用户影响力动态评价模型.该模型对用户影响力进行多层次分解,建立模糊情境下的指标体系,以区间数描述用户数据,同时引入时间维度考察数据的动态差异,采用新的熵方法计算模糊熵,设计主客观相结合的熵权确定方法,从而对用户影响力进行量化分析.克服了主观赋权的局限,提供了社交网络量化评价的新思路,拓展了区间直觉模糊群决策方法的应用.最后,应用该模型对新浪微博用户进行影响力动态评价,验证模型的有效性.A new entropy measure of interval-valued intuitionistic fuzzy set is proposed based on several limitations in the existing methods. Considering the highly dynamic and unstructured characteristics of social network data, we innovatively take the influence measure of social network users as a fllzzy group decision-making issue with the interval-valued intuitionistic fuzzy thought, thus put forward a user influence dynamic evaluation model. This new model decomposes user influences nmlti-levelly, sets up an index system in fuzzy situation, uses interval-valued intuitionistic fuzzy numbers for data description, meanwhile, considers the differences in time dimension, uses the new entropy measure for calculation and then designs a entropy weight method which combines subjectivity and objectivity together, so as to analyze user influences quantitatively. This model offers a new thinking of social network quantitative evaluation, expands the application range of interval-valued intuitionistie fuzzy decision-making approach. Finally, we apply this model for influence dynamic evaluation of Sina Weibo users to prove its validity.

关 键 词：多属性群决策 区间直觉模糊熵 熵权法 社交网络 微博 用户影响力 

分 类 号：F49[经济管理—产业经济] TP393.092[自动化与计算机技术—计算机应用技术]