Trust in AI systems for project and risk management: evaluating the role of transparency, reputation, technical competence, and reliability
Trust in AI systems for project and risk management: evaluating the role of transparency, reputation, technical competence, and reliability
Machine learning algorithms are often perceived as opaque, undermining user trust in AI systems. Explainable AI (XAI) seeks to mitigate this by enhancing transparency through clear explanations of algorithmic predictions. Furthermore, the reliability of these systems can be improved by integrating predictions from multiple algorithms. This study developed a hybrid project prediction system that merges XAI with several machine learning algorithms, thus fostering trust among project professionals and decision-makers. The theoretical model, grounded in literature on technology adoption, inter-organisational relationships, and XAI, examines four key trust factors: transparency, reputation, technical competence, and reliability. Employing a survey experiment and structural equation modelling, the research provides a nuanced understanding of how these factors influence trust in AI applications within project and risk management, contributing significantly to both academic literature and practical implementations.
Trustworthiness, AI, Explainable AI, Project Management, Machine Learning, Transparency
Hsu, Ming-Wei
1321a3d0-e965-4438-b981-95aba5d0394c
Dacre, Nicholas
90ea8d3e-d0b1-4a5a-bead-f95ab32afbd1
Senyo, P.K.
a997594f-9a9f-411a-a097-bac972452c6f
12 September 2023
Hsu, Ming-Wei
1321a3d0-e965-4438-b981-95aba5d0394c
Dacre, Nicholas
90ea8d3e-d0b1-4a5a-bead-f95ab32afbd1
Senyo, P.K.
a997594f-9a9f-411a-a097-bac972452c6f
Hsu, Ming-Wei, Dacre, Nicholas and Senyo, P.K.
(2023)
Trust in AI systems for project and risk management: evaluating the role of transparency, reputation, technical competence, and reliability.
In Operational Research Society.
Record type:
Conference or Workshop Item
(Paper)
Abstract
Machine learning algorithms are often perceived as opaque, undermining user trust in AI systems. Explainable AI (XAI) seeks to mitigate this by enhancing transparency through clear explanations of algorithmic predictions. Furthermore, the reliability of these systems can be improved by integrating predictions from multiple algorithms. This study developed a hybrid project prediction system that merges XAI with several machine learning algorithms, thus fostering trust among project professionals and decision-makers. The theoretical model, grounded in literature on technology adoption, inter-organisational relationships, and XAI, examines four key trust factors: transparency, reputation, technical competence, and reliability. Employing a survey experiment and structural equation modelling, the research provides a nuanced understanding of how these factors influence trust in AI applications within project and risk management, contributing significantly to both academic literature and practical implementations.
This record has no associated files available for download.
More information
Published date: 12 September 2023
Additional Information:
This study is particularly relevant for project professionals and researchers interested in integrating AI within project and risk management frameworks. It sheds light on how trust in autonomous systems can be enhanced through factors such as transparency, technical competence, reputation, and reliability. Employing a hybrid model that combines Explainable AI (XAI) with multiple machine learning algorithms, the research offers a theoretical model for evaluating essential trust factors. These insights are salient for informed decision-making and deepen understanding of the strategic applications of AI technologies in project management.
Keywords:
Trustworthiness, AI, Explainable AI, Project Management, Machine Learning, Transparency
Identifiers
Local EPrints ID: 492469
URI: http://eprints.soton.ac.uk/id/eprint/492469
PURE UUID: 371ea666-9a41-4aac-97e9-d7f383a64c8e
Catalogue record
Date deposited: 29 Jul 2024 16:57
Last modified: 08 Nov 2024 02:56
Export record
Contributors
Author:
Ming-Wei Hsu
Author:
P.K. Senyo
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
