A two-sided matching model in the context of B2B export cross-border e-commerce

Abstract

Cross-border electronic commerce plays an increasingly key role in international trades, which has become the focus of concern in both academia and industry. However, how to better match overseas demanders and domestic suppliers is still a question for business-to-business export agent. To achieve a steady state, in this study, we apply the two-sided matching method to business-to-business export cross-border electronic commerce context based on the satisfaction of different stakeholders, i.e., sellers, buyers, platforms, and third-party service providers. The satisfaction degree is based on a series of indicators extracted from the prior literature. By incorporating both the linguistic and interval assessment information, we theoretically build an optimization model to match the two sides: overseas demanders and domestic suppliers. Then we empirically illustrate the feasibility and performance of the model through the numerical simulations. The operation mechanisms and managerial guidelines for the two-sided matching in business-to-business export cross-border e-commerce are presented. Our study makes contributions in both theory and practice.

Supply and demand matching model of P2P sharing accommodation platforms considering fairness

“…We designed a calculation method for the proportion of similar matching pairs (similarity rate) between two different data sets and the average similarity rate of matching results between these data sets, and expressed with Eqs. (34) and (35). In the equation, NS ij and NT i represent the number of similar matching pairs between dataseti and datasetj, and the number of matching pairs of each data set, respectively.…”

“…Based on Eqs. (34) and (35), we design an algorithm to calculate the similarity rate of matching results (Algorithm 1). (34)…”

“…Fairness of matching refers to the preferences of both sides are treated fairly in the matching process, and the matching goal is the minimization of differences between these preferences [ 30 , 31 ], to ensure that the preferences of two sides are as close as possible and reduce differences and jealousy, which is always used as a support for the matching stability. For better matching results, the matching optimization strategies with fairness increase gradually, such as issue of matching on uncertain preference sequence [ 32 ], supply meet demand of technological knowledge [ 33 ], task match resource in cloud manufacturing [ 34 ], selection of foreign customers in B2B export cross-border e-commerce [ 35 ]. They achieve the fairness of matching mainly by setting weight coefficients for the preferences of both parties, such as (0.5, 0.5) [ 36 , 37 ], or the minimization of the absolute value of preference difference [ 32 , 34 ].…”

“…Preferences information is the key basis for matching decisions and often used in the research and practice of matching decision-making, such as cross-border e-commerce [ 35 ], ridesharing system [ 37 ], cloud manufacturing [ 34 , 46 ], and smart intelligent technique transfer [ 47 ]. Integrating the extant studies on preferences and satisfaction in e-trading and two-sided matching [ 34 , 37 , 48 , 49 ], we advanced the definition of preferences of P2P platform peers as follows.…”

“…In order to consider the consistency of preferences in the matching process, the preferences balance is achieved by assigning the same weight (e.g. 0.5 and 0.5) to both sides in [ 36 , 37 ], or by minimizing the difference between the two preferences as in [ 32 , 34 ], or the square root of the product of the two preferences in [ 35 ]. However, the preferences of P2P platform supply and demand sides are diversified and heterogeneous, so it is difficult to deal with them with a unified standard.…”