A fair, preference-based posted price resale e-market model and clearing heuristics for circular economy

Özer A. H.

Applied Soft Computing, vol.106, 2021 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 106
  • Publication Date: 2021
  • Doi Number: 10.1016/j.asoc.2021.107308
  • Title of Journal : Applied Soft Computing
  • Keywords: Resale market, Circular economy, Ant colony optimization, Artificial bee colony, Genetic algorithm, STRUCTURAL MODEL, COLONY, OPTIMIZATION, AUCTION, ALGORITHM


© 2021 Elsevier B.V.Resale markets in which secondhand new and used goods are traded play an important role in circular economy with significant economic and environmental benefits. This study proposes a preference-based posted price electronic market model for resale markets which features several mechanisms to improve the market outcome. The proposed model allows market participants to post sales and purchase orders simultaneously inside a trading round which also enables participants to use the revenue obtained from the items to be sold to purchase other items in the market. Besides, each participant is allowed to declare a budget constraint which restricts the amount that the participant will spend in the market to prevent a possible budget deficiency. Furthermore, the model also allows participants to declare their preferences of substitutabilities in their orders. In this study, the proposed model is formally defined, the corresponding market clearing problem is formulated as a hierarchical multi-objective linear integer program to provide fair allocation between the participants. Four different objective functions are proposed, and their outcomes are compared to the current market system. Since the clearing problem is NP-Hard, several heuristic methods including ant colony optimization, artificial bee colony and genetic algorithms along with problem-specific operators are proposed. The performance of the model is statistically analyzed based on several experiments. The genetic algorithm using the proposed problem-specific operators provides solutions within 3% of the optimal objective values and within 1% of the optimal fairness on average. The results also indicate that the model provides improved market outcomes and fair allocation of items among the participants, and thus it has a potential to contribute to the growth of circular economy.