Linguistic fuzzy consensus model for collaborative development of fuzzy cognitive maps: a case study in software development risks


          

刊名：Fuzzy Optimization and Decision Making: A Journal of Modeling and Computation Under Uncertainty
作者：De Maio, Carmen（Univ Salerno, Dipartimento Ingn Informaz Elettr & Matemat Appli, I-84084 Fisciano, SA, Italy）
Fenza, Giuseppe（Univ Salerno, Dept Management & Innovat Syst, I-84084 Fisciano, SA, Italy）
Loia, Vincenzo（Univ Salerno, Dept Management & Innovat Syst, I-84084 Fisciano, SA, Italy）
Orciuoli, Francesco（Univ Salerno, Dept Management & Innovat Syst, I-84084 Fisciano, SA, Italy）
刊号：519LB020
ISSN：1568-4539
出版年：2017
年卷期：2017, vol.16, no.4
页码：463-479
总页数：17
分类号：O22
关键词：Group decision making；Fuzzy consensus model；Fuzzy cognitive maps
参考中译：
语种：eng
文摘：Fuzzy cognitive maps (FCMs) have been widely used in several domains for information processing, such as, data fusion, decision making. Although several methods to automatically learn FCMs are recognized from the scientific literature, the most used approach to build an FCM relies on a collaborative task involving single person or, more suitably, group of experts. Collaborative development increases reliability and robustness of the resulting FCM, but rises some problems in terms of group decision making to aggregate different perspectives of the problem representation. This paper proposes to support collaborative development of FCMs introducing knowledge engineering process that relies on Linguistic Fuzzy Consensus Model. In the proposed approach, each expert builds the own version of the FCM. When all different versions are available, a Group Decision Making process is activated in order to reach the consensus on conflictual modeling opinions. The result is a unique final version of the FCM that is not a simple aggregation of the versions provided by the experts but is the result of a well-suited mathematical model. In addition, this work adopts consensus model with incomplete preference relations scheme to address knowledge harmonization issues. Finally, advantages and the limitations of the proposed framework are argued.



©2016机械工业出版社（机械工业信息研究院） 京ICP备05055788号-35