A new multi-granularity probabilistic linguistic two-sided matching method considering peer effect and its application in pension services

• Published in: International journal of machine learning and cybernetics Vol. 13; no. 7; pp. 1907 - 1926
• Main Authors: Wang, Nannan, Li, Peng
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.07.2022; Springer Nature B.V
• Subjects: Aging; Artificial Intelligence; Complex Systems; Computational Intelligence; Control; Decision making; Engineering; Interpersonal relations; Linguistics; Matching; Mechatronics; Older people; Original Article; Pattern Recognition; Robotics; Social networks; Suppliers; Supply & demand; Systems Biology; China; Multi-granularity probabilistic linguistic term sets; Peer effect; Decision making; Two-sided matching; Pension services
• ISSN: 1868-8071; 1868-808X
• DOI: 10.1007/s13042-021-01495-2

China is now facing a serious aging trend. With the increase in the demand for pension, the pension mode tends to be diversified. On one hand, the elderly want to select suitable pension modes based on their family situations. On the other hand, different pension modes should choose applicable elderly people according to their service characteristics. This two-way selection can be seen as a two-sided matching problem. Because this matching environment is usually complex and uncertain, multi-granularity probabilistic linguistic term set (PLTS) is an effective tool to describe the uncertain evaluation process. Furthermore, complex social relations of decision makers will affect decision-making results. Therefore, this paper aims to provide a two-sided matching method based on the peer effect using multi-granularity PLTSs, which can handle uncertain evaluation information accurately considering influence of social network relationships. Firstly, we propose a conversion method of multi-granularity PLTSs based on two-tuple linguistics. Then, we put forward a calculation formula for matching satisfaction with PLTSs with the consideration of the expectations and sensitivity to satisfaction for subjects. Furthermore, a multi-objective two-sided matching model with maximum satisfaction is established. Finally, we apply our method to a real case for matching of pension services and make comparisons with a traditional two-sided method.