An explainable decision intelligence framework for consumer packaging evaluation in e-commerce: a hybrid entropy-weighted grey MCDM approach
Year: 2026
Type: article
Abstract: With the rapid application of transformative technologies such as federated learning and generative AI in consumer terminals and digital commerce, there is a growing demand for decision-support frameworks that are intelligent, interpretable, and privacy-protective.In the e-commerce field, consumer satisfaction is increasingly dependent on data-driven product evaluation mechanisms rather than isolated design attributes.This study proposes a grey multi-criteria decision-making (G-MCDM) framework integrated with entropy weighting to evaluate consumer satisfaction with various types of cosmetic packaging, including bottles, jars, hoses, and aerosol cans.The model integrates entropy-based objective weighting methods and grey system theory, effectively addressing the uncertainty, heterogeneity, and information incompleteness present in large-scale consumer feedback data.Empirical analysis based on e-commerce platform data shows that tube packaging achieves the highest overall consumer satisfaction among all tested packaging forms.Beyond the specific application scenario, this framework can serve as an interpretable decision-making module, embedded in the federated learning architecture to support decentralised consumer data analysis and generative AI-assisted packaging design optimisation.This research improves the transparent and scalable evaluation mechanism suitable for AI-enabled consumer scenarios, and clarifies the core role of hybrid decision intelligence models in supporting the next-generation digital e-commerce ecosystem.
Authors: Wenchao Pan, Lianhua Liu, Peng Qiang, Fangting Zhang, Luanling Tan, Xingyu Zou, Xiao Wang
Institution: Macao Polytechnic University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W7203675809