Definition of a DSL focused on Integrating Machine Learning into IoT-Enhanced BPs

The integration of Machine Learning into IoT-Enhanced Business Processes (BPs) offers significant opportunities to enable predictive and adaptive process behavior. However, in current practice, ML capabilities are typically developed through ad-hoc pipelines that remain disconnected from process mod...

Descripción completa

Detalles Bibliográficos
Autores: Valderas, Pedro|||0000-0002-4156-0675, Bouich, Asmaa
Tipo de recurso: informe técnico
Fecha de publicación:2026
País:España
Institución:Universitat Politècnica de València (UPV)
Repositorio:RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Idioma:inglés
OAI Identifier:oai:dnet:riunet______::92814a347e27a917d1245eb88e44cd2b
Acceso en línea:https://riunet.upv.es/handle/10251/234384
Access Level:acceso abierto
Palabra clave:BPMN
Business Process
IoT
Machine Learning
Descripción
Sumario:The integration of Machine Learning into IoT-Enhanced Business Processes (BPs) offers significant opportunities to enable predictive and adaptive process behavior. However, in current practice, ML capabilities are typically developed through ad-hoc pipelines that remain disconnected from process modeling and execution, leading to limited traceability, maintainability, and reuse. To address this challenge, this paper proposes a Domain-Specific Language (DSL) that supports the systematic and model-driven integration of ML into IoT-Enhanced BPs. The DSL enables the explicit specification of heterogeneous data sources, feature engineering operations, ML-ready datasets, and ML model configurations within a unified and platform-independent framework. The approach is supported by a model-to-code transformation that generates executable ML pipelines.