A value-based approach to agent-based simulation for policy assessment: an exploration in the water domain

• Autores: Antonio Perelló
• Directores de la Tesis: Pablo Cayetano Noriega Blanco-Vigil (dir. tes.)
• Lectura: En la Universitat Autònoma de Barcelona ( España ) en 2020
• Idioma: español
• Tribunal Calificador de la Tesis: Néstor Duch Brow (presid.), Enric Plaza Cervera (secret.), Wernher Marco Schorlemmer (voc.), Hèctor Monclús Sales (voc.), Joaquim Comas Matas (voc.)
• Programa de doctorado: Programa Oficial de Doctorado en Informática
• Materias:
  • Matemáticas
    • Ciencia de los ordenadores
      • Inteligencia artificial
      • Simulación
  • Sociología
    • Sociología matemática
      • Elaboración de modelos
• Texto completo no disponible (Saber más ...)
• Resumen

  • Policy-making is a value-driven activity. From the assessment of states of the world to the choice of the appropriate actions to do under given circumstances, values are present in deciding whether something is good or not. Thus, values are involved in policy design, but also in the behaviour of those individuals who are going to be affected by a policy. Since these value-driven choices in policy-making may have significant social consequences, it is advisory to assess policies prior to their enactment. Assuming that agent-based simulation is a powerful methodology for this purpose, the need of a conceptual framework that includes values into policy simulator systems is clear. Interestingly, this leads to explore the role of values in problem design from the ground, especially the model and the outcome assessment. In order to imbue values into computational models, a methodology is proposed: contextualise and translate abstract values into state indicators, and build the simulator around these representations (from the set of interventions to the their assessment). This methodology proposal is illustrated in the water domain, and is used to address two case studies by means of simulation: the first, the modernisation of farmer communities in Spain; and second, the interaction of policy actors that lead to management policy shifts in the urban context. Finally, the contributions of the framework in issues that are raising concerns in Artificial Intelligence nowadays, as well as potential contributions to water management, are explained as future work lines.