Trustworthy human-centric based Automated Decision-Making Systems
Authors:
Marcelino Cabrera,
Carlos Cruz,
Pavel Novoa-Hernández,
David A. Pelta,
José Luis Verdegay
Abstract:
Automated Decision-Making Systems (ADS) have become pervasive across various fields, activities, and occupations, to enhance performance. However, this widespread adoption introduces potential risks, including the misuse of ADS. Such misuse may manifest when ADS is employed in situations where it is unnecessary or when essential requirements, conditions, and terms are overlooked, leading to uninte…
▽ More
Automated Decision-Making Systems (ADS) have become pervasive across various fields, activities, and occupations, to enhance performance. However, this widespread adoption introduces potential risks, including the misuse of ADS. Such misuse may manifest when ADS is employed in situations where it is unnecessary or when essential requirements, conditions, and terms are overlooked, leading to unintended consequences. This research paper presents a thorough examination of the implications, distinctions, and ethical considerations associated with digitalization, digital transformation, and the utilization of ADS in contemporary society and future contexts. Emphasis is placed on the imperative need for regulation, transparency, and ethical conduct in the deployment of ADS.
△ Less
Submitted 22 December, 2023;
originally announced January 2024.
Trust dynamics and user attitudes on recommendation errors: preliminary results
Authors:
David A. Pelta,
Jose L. Verdegay,
Maria T. Lamata,
Carlos Cruz Corona
Abstract:
Artificial Intelligence based systems may be used as digital nudging techniques that can steer or coerce users to make decisions not always aligned with their true interests. When such systems properly address the issues of Fairness, Accountability, Transparency, and Ethics, then the trust of the user in the system would just depend on the system's output. The aim of this paper is to propose a mod…
▽ More
Artificial Intelligence based systems may be used as digital nudging techniques that can steer or coerce users to make decisions not always aligned with their true interests. When such systems properly address the issues of Fairness, Accountability, Transparency, and Ethics, then the trust of the user in the system would just depend on the system's output. The aim of this paper is to propose a model for exploring how good and bad recommendations affect the overall trust in an idealized recommender system that issues recommendations over a resource with limited capacity. The impact of different users attitudes on trust dynamics is also considered. Using simulations, we ran a large set of experiments that allowed to observe that: 1) under certain circumstances, all the users ended accepting the recommendations; and 2) the user attitude (controlled by a single parameter balancing the gain/loss of trust after a good/bad recommendation) has a great impact in the trust dynamics.
△ Less
Submitted 11 February, 2020;
originally announced February 2020.
