Explainable AI for customer segmentation in product development

AI can facilitate customer segmentation in product development. However, the black-box nature of AI often causes designers to distrust AI predictions. Explainable AI (XAI) is an emerging AI paradigm that can provide humanly understandable explanations about AI predictions. The outputs of XAI, feature-based and data-based explanations, can improve AI performance and designer trust in AI. A new framework is proposed to incorporate XAI explanations into customer segmentation. An experiment is conducted to validate the framework effectiveness. It indicates that AI performance can be enhanced by XAI, where feature-based explanations can facilitate feature selection and data-based explanations can uncover high-value datasets.