A Unified Fuzzy–Explainable AI Framework (FAS-XAI) for Customer Service Value Prediction and Strategic Decision-Making
This article introduces FAS-XAI, a unified and interpretable methodological framework for decision-making under uncertainty, integrating fuzzy clustering and explainable artificial intelligence (XAI). The framework combines three core components: Fuzzy C-Means (FCM) to identify overlapping behaviora...
| Autor: | |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2025 |
| País: | España |
| Institución: | Universidad Europea (UEM) |
| Repositorio: | ABACUS. Repositorio de Producción Científica |
| Idioma: | inglés |
| OAI Identifier: | oai:abacus.universidadeuropea.com:11268/16671 |
| Acceso en línea: | https://hdl.handle.net/11268/16671 |
| Access Level: | acceso abierto |
| Palabra clave: | Economía Informática Inteligencia artificial Goal 8: Promote inclusive and sustainable economic growth, employment and decent work for all Goal 9: Build resilient infrastructure, promote sustainable industrialization and foster innovation Goal 12: Ensure sustainable consumption and production patterns |
| Sumario: | This article introduces FAS-XAI, a unified and interpretable methodological framework for decision-making under uncertainty, integrating fuzzy clustering and explainable artificial intelligence (XAI). The framework combines three core components: Fuzzy C-Means (FCM) to identify overlapping behavioral profiles in ambiguous contexts, supervised machine learning models to predict decision outcomes, and expert-guided interpretation to contextualize results and enhance transparency. Global and local explainability is ensured through SHAP, LIME, and ELI5, placing human reasoning at the center of the decision process. The methodology is validated using a real-world B2B customer service dataset from a global ERP software distributor, where customer engagement is modeled through the RFID approach (Recency, Frequency, Importance, Duration). Fuzzy clustering uncovers latent customer profiles, while XGBoost models predict attrition risk with interpretable explanations. The case study demonstrates the coherence, transparency, and operational value of FAS-XAI for strategic customer relationship management, and the framework is further discussed as a general-purpose, human-centered approach applicable across domains such as education, physics, and industry. |
|---|