Some HCI Priorities for GDPR-Compliant Machine Learning

Veale, Michael; Binns, Reuben; Kleek, Max Van

doi:10.31228/osf.io/wm6yk

Cited by 7 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…AI and ML may be the tools used to provide such automated decision making, and, therefore, it places these technologies in a legal spotlight. Some guidelines for GDPR-compliant ML development have recently been provided by Veale et al [7].…”

Section: Societal Issues Of Ai and ML Applicationmentioning

confidence: 99%

Societal Issues Concerning the Application of Artificial Intelligence in Medicine

Vellido

2018

Kidney Dis

View full text Add to dashboard Cite

Background: Medicine is becoming an increasingly data-centred discipline and, beyond classical statistical approaches, artificial intelligence (AI) and, in particular, machine learning (ML) are attracting much interest for the analysis of medical data. It has been argued that AI is experiencing a fast process of commodification. This characterization correctly reflects the current process of industrialization of AI and its reach into society. Therefore, societal issues related to the use of AI and ML should not be ignored any longer and certainly not in the medical domain. These societal issues may take many forms, but they all entail the design of models from a human-centred perspective, incorporating human-relevant requirements and constraints. In this brief paper, we discuss a number of specific issues affecting the use of AI and ML in medicine, such as fairness, privacy and anonymity, explainability and interpretability, but also some broader societal issues, such as ethics and legislation. We reckon that all of these are relevant aspects to consider in order to achieve the objective of fostering acceptance of AI- and ML-based technologies, as well as to comply with an evolving legislation concerning the impact of digital technologies on ethically and privacy sensitive matters. Our specific goal here is to reflect on how all these topics affect medical applications of AI and ML. This paper includes some of the contents of the “2nd Meeting of Science and Dialysis: Artificial Intelligence,” organized in the Bellvitge University Hospital, Barcelona, Spain. Summary and Key Messages: AI and ML are attracting much interest from the medical community as key approaches to knowledge extraction from data. These approaches are increasingly colonizing ambits of social impact, such as medicine and healthcare. Issues of social relevance with an impact on medicine and healthcare include (although they are not limited to) fairness, explainability, privacy, ethics and legislation.

show abstract

Section: Societal Issues Of Ai and ML Applicationmentioning

confidence: 99%

Societal Issues Concerning the Application of Artificial Intelligence in Medicine

Vellido

2018

Kidney Dis

View full text Add to dashboard Cite

show abstract

“…This can be especially challenging, since regional legal requirements can differ substantially, i.e. GDPR in Europe in contrast to US legislation (Veale et al , 2018). Thus, a company may need to adapt its approach based on the country it is operating in.…”

Section: Discussionmentioning

confidence: 99%

Will the customers be happy? Identifying unsatisfied customers from service encounter data

Baier

Kühl

Schüritz

et al. 2020

JOSM

View full text Add to dashboard Cite

PurposeWhile the understanding of customer satisfaction is a key success factor for service enterprises, existing elicitation approaches suffer from several drawbacks such as high manual effort or delayed availability. However, the rise of analytical methods allows for the automatic and instant analysis of encounter data captured during service delivery in order to identify unsatisfied customers.Design/methodology/approachBased on encounter data of 1,584 IT incidents in a real-world service use case, supervised machine learning models to predict unsatisfied customers are trained and evaluated.FindingsWe show that the identification of unsatisfied customers from encounter data is well feasible: via a logistic regression approach, we predict dissatisfied customers already with decent accuracy—a substantial improvement to the current situation of “flying blind”. In addition, we are able to quantify the impacts of key service elements on customer satisfaction.Research limitations/implicationsThe possibility to understand the relationship between encounter data and customer satisfaction will offer ample opportunities to evaluate and expand existing service management theories.Practical implicationsIdentifying dissatisfied customers from encounter data adds a valuable methodology to customer service management. Detecting unsatisfied customers already during the service encounter enables service providers to immediately address service failures, start recovery actions early and, thus, reduce customer attrition. In addition, providers will gain a deeper understanding of the relevant drivers of customer satisfaction informing future new service development.Originality/valueThis article proposes an extendable data-based approach to predict customer satisfaction in an automated, timely and cost-effective way. With increasing data availability, such AI-based approaches will spread quickly and unlock potential to gain important insights for service management.

show abstract

“…Moreover, Article 25 requires "Data protection by design and by default" from all data controllers, regardless of their size. The controllers must be able to demonstrate that they have implemented dedicated measures to protect these principles and that they have integrated specific state-of-the-art safeguards that are necessary to secure rights and freedoms of data subjects [55]. Therefore, medical professionals using this software must consequently have an understanding of the mechanisms underlying their function [56,57].…”

Section: Limits Of Aimentioning

confidence: 99%

Artificial intelligence in thoracic surgery: past, present, perspective and limits

et al. 2020

View full text Add to dashboard Cite

Artificial intelligence (AI) technology is becoming prevalent in many areas of everyday life. The healthcare industry is concerned by it even though its widespread use is still limited. Thoracic surgeons should be aware of the new opportunities that could affect their daily practice, by direct use of AI technology or indirect use via related medical fields (radiology, pathology and respiratory medicine). The objective of this article is to review applications of AI related to thoracic surgery and discuss the limits of its application in the European Union. Key aspects of AI will be developed through clinical pathways, beginning with diagnostics for lung cancer, a prognostic-aided programme for decision making, then robotic surgery, and finishing with the limitations of AI, the legal and ethical issues relevant to medicine. It is important for physicians and surgeons to have a basic knowledge of AI to understand how it impacts healthcare, and to consider ways in which they may interact with this technology. Indeed, synergy across related medical specialties and synergistic relationships between machines and surgeons will likely accelerate the capabilities of AI in augmenting surgical care.

show abstract

Some HCI Priorities for GDPR-Compliant Machine Learning

Cited by 7 publications

References 27 publications

Societal Issues Concerning the Application of Artificial Intelligence in Medicine

Societal Issues Concerning the Application of Artificial Intelligence in Medicine

Will the customers be happy? Identifying unsatisfied customers from service encounter data

Artificial intelligence in thoracic surgery: past, present, perspective and limits

Contact Info

Product

Resources

About