<p>This study explores a clustering method that groups objects based on their incidence in a large number of given sets, with the objective of minimizing the occurrence of multiple objects from the same cluster in the same set. The computational aspects of the method are the main focus of the study. First, it is proven that the problem of finding the optimal clustering is NP-hard. Second, a genetic algorithm that is augmented by a renumbering procedure, a fast task-specific local search heuristic, and an initial solution based on a simplified model is proposed to numerically find a suitable clustering using k-means algorithm. Third, it is demonstrated in a simulation study that the improvements of the standard genetic algorithm significantly enhance its computational performance. Fourth, the usability of the proposed approach is illustrated in an application using the retail store data.</p>