A trust-similarity analysis-based clustering method for large-scale group decision-making under a social network
| Autor: | Zhi-jiao Du, Han-yang Luo, Su-min Yu, Xu-dong Lin |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Social network
business.industry Computer science Decision theory 020206 networking & telecommunications Context (language use) 02 engineering and technology computer.software_genre Field (computer science) Plot (graphics) Group decision-making Similarity (network science) Hardware and Architecture Signal Processing 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Data mining business Cluster analysis computer Software Information Systems |
| Zdroj: | Information Fusion. 63:13-29 |
| ISSN: | 1566-2535 |
| DOI: | 10.1016/j.inffus.2020.05.004 |
| Popis: | Large-scale group decision-making (LSGDM) under a social network context has attracted much attention in the field of decision science. The clustering of individual opinions and the handling of trust relationships are the main research topics. Opinion similarity and trust relationship are considered to be two important measurement attributes for implementing clustering. Traditional clustering methods often use a single attribute to divide the original group without requiring a combination of the above two attributes. However, these two attributes play different roles in the clustering process, insofar as opinion similarity is used to measure the level of difference among individual opinions, whereas the trust relationship represents the trustworthiness of decision makers. This paper develops a trust-similarity analysis (TSA)-based clustering method to manage the clustering operation in LSGDM events under a social network context. First, the trust-similarity matrix is established to collectively describe the decision information. Second, all measurement attribute values are mapped to a trust-similarity plot from which the joint threshold can be calculated. Finally, a TSA-based clustering method is proposed that considers the attributes of opinion similarity and trust relationship and that allocates their importance to achieve specific clustering objectives. The numerical experiment and comparative analysis reveal the feasibility and advantages of the proposed method. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect