A new inclusion measure-based clustering method and its application to product classification

Appropriate product classification can not only promote the sale of products but also highly improve the profits of manufacturers and retailers. To derive a desirable clustering of products, we need to consider qualitative (e.g., customers' evaluation) and quantitative (e.g., product sales) information simultaneously. However, traditional methods such as K -means clustering, hierarchical clustering, and density-based clustering fail to depict the fuzzy cognitions of decision makers (DMs) in real cases. This article investigates clustering methods to aid DMs in their classification management of products. To do so, an intuition-istic multiplicative set (IMS), a new form of information expression, is applied to express the DMs' evaluations of products due to its strong ability to handle unbalanced or asym-metric information. This study includes an in-depth analysis of inclusion measures of IMSs to measure the inclusion relationships between products under the given evaluation criteria. We propose two types of clustering techniques-the intuitionistic multiplicative (IM) transitive closure clustering method and the IM netting clustering method-to tackle the clustering problem with IMS information. Lastly, a case study of product classification management is presented to illustrate the applicability of the proposed methods. Furthermore, the validity and superiority of these clustering methods are verified by the-oretical and practical comparisons.(c) 2023 Elsevier Inc. All rights reserved.