A sentiment analysis-based two-stage consensus model of large-scale group with core-periphery structure
A sentiment analysis-based two-stage consensus model of large-scale group with core-periphery structure
复制标题
基于情感分析的核心-外围结构大规模群体两阶段共识模型
DOI：
10.1016/j.ins.2022.11.147
复制
发表时间：
2022-12
影响因子：
8.1
通讯作者：
Aihua Wang
中科院分区：
文献类型：
--
作者：
Yuanyuan Liang;Yanbing Ju;Peiwu Dong;Xiao-Jun Zeng;Luis Martínez;Jinhua Dong;Aihua Wang
大数据和社交媒体的发展推动了大规模群体决策（LSGDM）与社交网络的融合，并关注个体行为因素。基于这一趋势，本文提出了一种新的LSGDM共识模型，该模型探索和管理了社会网络环境下专家使用自由文本表达观点的中尺度结构。在该方法中，首先采用情感分析来提取对专家提供的备选方案的偏好，并将偏好进一步转换为分布式语言偏好关系矩阵。在此基础上，提出了一种基于语言分布评价距离测度的社会网络核心-边缘检测方法。在此基础上，专家的权重是由一个优化模型，最大限度地提高专家的可靠性的基础上一致性和节点中心。考虑到被检测网络中成员的参考依赖性和有限理性特征，提出了一种基于前景理论的两阶段共识模型，以系统地、渐进地提高群体共识度。最后，一个关于生命科学投资的案例研究来说明我们的建议的有用性。理论和仿真分析证明了该模型的收敛性。对比分析揭示了该模型的特点和优势。
The development of big data and social media has driven large-scale group decision making (LSGDM) to merge with social networks and focus on individual behavioral factors. Following this trend, this paper develops a novel LSGDM consensus model that explores and manages themeso-scale structure among experts using free texts to express their opinions under social network settings. In the proposed approach, firstly the sentiment analysis is adopted to extract preferences over alternatives provided by experts and the preferences are further converted into distributed linguistic preference relation matrices. Then a core-periphery detection method for the social network constructed based on the newly defined distance measure for linguistic distribution assessments is proposed. After that, expert weights are derived by an optimization model that maximizes the expert reliability based on consistency and node centrality. Moreover, considering reference dependence and bounded rationality features of members among the detected network, a prospect theory-based two-stage consensus model is developed to improve group consensus systematically and gradually. Finally, a case study regarding life science investments is provided to illustrate the usefulness of our proposal. The convergence of the proposed model is proven by theoretical and simulation analysis. Comparative analysis reveals the features and advantages of our model.
登录
查看更多内容
一致性和共识驱动的模型，用于个性化语言术语的个体语义，以支持具有分布语言偏好关系的群体决策
影响因子：
8.8
作者：
Tang Xiaoan;Peng Zhanglin;Zhang Qiang;Pedrycz Witold;Yang Shanlin
通讯作者：
Yang Shanlin
DOI：
10.1017/cbo9780511609220.014
发表时间：
1979-12
期刊：
--
影响因子：
--
作者：
D. Kahneman;A. Tversky
通讯作者：
D. Kahneman;A. Tversky
DOI：
10.1016/j.inffus.2021.05.003
发表时间：
2021-12
期刊：
Inf. Fusion
影响因子：
--
作者：
Haiming Liang;Xia Chen;Congcong Li;Hengjie Zhang
通讯作者：
Haiming Liang;Xia Chen;Congcong Li;Hengjie Zhang
DOI：
10.1016/j.cie.2003.12.012
发表时间：
2004-04
期刊：
Comput. Ind. Eng.
影响因子：
--
作者：
Z. Fan;Si-Han Xiao;G. Hu
通讯作者：
Z. Fan;Si-Han Xiao;G. Hu
影响因子：
8.1
作者：
Liao, Huchang;Li, Xiaofang;Tang, Ming
通讯作者：
Tang, Ming