Abstract
This article considers how combinatorial computing can be used to improve intelligent group-buying systems. The main idea is quite simple: a buyer should not be forced to choose only one purchasing group if several offers are relevant to them. The system therefore searches for group combinations that give the best total benefit while still respecting budgets, product limits, and minimum group-size conditions. In the study, this task is written as an integer optimization problem and then tested on simulated market scenarios. The experiments vary preference diversity, the number of users, and the strictness of constraints. The obtained results show that the method can form feasible groups with reasonable efficiency and does not lose stability when the scenario becomes more difficult. This makes the approach useful for group-buying platforms that need to describe consumer behavior in a more realistic way.