Skip to main content

INTEGRACION DE IOT CON PARTICIPACION HUMANA Y TECNOLOGIAS DE DIGITAL TWINS PARA LA MEJORA DE PROCESOS EN LA ERA DE LA INTELIGENCIA HIBRIDA

PID2023-146224OB-I00

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprogram Subprograma Estatal de Generación de Conocimiento
Call Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Call year 2023
State Plan 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 1
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Contribución de congreso (ConferenceOutput)

Explainability Across the Spectrum: Modeling Stakeholder Goals Based on AI Complexity Levels

  • Antoni Mestre
  • Albert Albiol, Manuela|||0000-0003-3747-400X
  • Pelechano Ferragud, Vicente|||0000-0003-1090-230X
  • Gil, Miriam
[EN] Explainability is a critical enabler of trust, usability, and regulatory compliance in artificial intelligence (AI) systems. In Hybrid Intelligence (HI) systems, where human expertise and AI collaborate to make decisions, the need for explainability becomes even more pronounced. However, the complexity of explainability requirements varies significantly depending on the role AI plays within the system. From human-dominant systems, where AI provides supportive insights, to AI-dominant systems, where AI assumes autonomous decision-making, the nature and depth of explanations required shift across this spectrum. Despite its importance, there is a lack of systematic approaches for tailoring explainability to meet stakeholder needs across these varying levels of AI complexity. This paper identifies the key components that shape explainability in HI systems, addressing the distinct roles and interactions of human and AI agents, as well as the diverse needs of stakeholders. In addition, the paper introduces a conceptual framework to assist designers in integrating explain-ability as a Non-Functional Requirement into HI systems. The framework leverages Goal-Oriented Requirements Engineering to facilitate the incorporation of suitable explainability NFRs into the design of HI systems, enabling designers to address challenges such as enhancing transparency, fostering trust, and meeting diverse stakeholder expectations. This research provides a first approach for advancing the design of adaptable, accountable, and trustworthy HI systems, contributing valuable insights to the fields of requirements engineering, human-AI collaboration, and regulatory compliance., This work has been developed with the financial support of the of the Generalitat Valenciana with TENTACLE (CIAICO/2023/089) and by the MINECO under project PRODIGIOUS PID2023-146224OB-I00.