Scenario-Based Requirements Elicitation for User-Centric Explainable AI

Cirqueira, Douglas; Nedbal, Dietmar; Helfert, Markus; Bezbradica, Marija

doi:10.1007/978-3-030-57321-8_18

Cited by 30 publications

(34 citation statements)

References 89 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study conducted by [89] enforces the need to understand the operational environment of stakeholders who are going to use the explanations generated by XAI methods, to support their decision-making process. Author proposes a scenario-based requirement elicitations method for developing user-centric explanations using XAI methods for fraud detections.…”

Section: A Year-wise ML Methods Published In Aml Domain B Interpretability Of Models Used In Aml Solutions C Machine Learning Techniques mentioning

confidence: 99%

Deep Learning and Explainable Artificial Intelligence Techniques Applied for Detecting Money Laundering–A Critical Review

et al. 2021

View full text Add to dashboard Cite

Money laundering has been a global issue for decades, which is one of the major threat for economy and society. Government, regulatory and financial institutions are combating it together in their respective capacity, however still billions of dollars in fines by authorities make the headlines in the news. High-speed internet services have enabled financial institutions to deliver better customer experience through multi-channel engagements, which has led to exponential growth in transactions and new avenues for laundering the money for fraudsters. Literature shows the usage of statistical methods, data mining and Machine Learning (ML) techniques for money laundering detection, but limited research on Deep Learning (DL) techniques, primarily due to lack of model interpretability and explainability of the decisions made. Several studies are conducted on application of ML for Anti-Money Laundering (AML), and Explainable Artificial Intelligence (XAI) techniques in general, but lacks the study on usage of DL techniques together with XAI. This paper aims to review the current state-of-the-art literature on DL together with XAI for identifying suspicious money laundering transactions and identify future research areas. Key findings of the review are, researchers have preferred variants of Convolutional Neural Networks, and AutoEncoder; graph deep learning together with natural language processing is emerging as an important technology for AML; XAI use is not seen in AML domain; 51% ML methods used in AML are non-interpretable, 58% studies used sample of old real data; key challenges for researchers are access to recent real transaction data and scarcity of labelled training data; and data being highly imbalanced. Future research directions are, application of XAI techniques to bring-out explainability, graph deep learning using natural language processing (NLP), unsupervised and reinforcement learning to handle lack of labelled data; and joint research programs between research community and industry to benefit from domain knowledge and controlled access to data.

show abstract

Section: A Year-wise ML Methods Published In Aml Domain B Interpretability Of Models Used In Aml Solutions C Machine Learning Techniques mentioning

confidence: 99%

Deep Learning and Explainable Artificial Intelligence Techniques Applied for Detecting Money Laundering–A Critical Review

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Regarding the first set of constructs for artefact development, we rely on our previous study results to elicit 13 fraud expert's tasks when analyzing suspicious fraud cases [51]. In the referred study, we adopt expert interviews with a scenario-based method, and a systematic literature review [11]. Scenario-based elicitation facilitates an HCI and problem-centered perspective to identify stakeholder requirements, goals, tasks and knowledge to develop decision support systems [64].…”

Section: Methodsmentioning

confidence: 99%

“…However, they would like to have included examples of cases and fraudulent customer journeys. We suggested the addition of scenarios, such as developed in [11], and the experts agree those are good examples of additions to the DP for fulfilling that need. They also agree that the DP give a sense of better structuring practices and can serve as documentation to rely on for using explanation methods.…”

Section: Information Quality Of Design Principles With Syntactic and Semantic Criteriamentioning

confidence: 97%

“…EM methods fulfilling this DP are among the most used in extant Explainable AI studies, as the understanding of important features enables not only explanations for the AI behaviour but also to clarify if the model is working correctly, which is valuable for AI engineers aiming to optimize their models [29]. In the scenario of fraud detection, the analysis of important features is among the first tasks performed by experts when reviewing cases [11]. Therefore, without DP 3, experts would lack explanations for understanding whether the AI is focusing on the correct parameters, according to their domain knowledge in fraud cases.…”

Section: Dp3 Dp 3 Instantiation Enables Experts To Look Closely Into What Data Featuresmentioning

confidence: 99%

“…Requirements are not elucidated before the deployment of explanations, and they also lack prescriptive principles for the alignment between EM and fraud experts requirements [3,61]. Therefore, this can cause the lack of trust from fraud experts in AI predictions [11]. Since XAI research has a major focus on optimizing EM and AI models, and previous XAI studies for fraud lack a usercentric perspective, in this work, we develop design principles (DP) to prescribe the alignment of fraud expert's tasks to EM for enabling Explainable AI decision support (XAIDSS) in fraud detection.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards Design Principles for User-Centric Explainable AI in Fraud Detection

Cirqueira

Helfert

Bezbradica

2021

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Experts rely on fraud detection and decision support systems to analyze fraud cases, a growing problem in digital retailing and banking. With the advent of Artificial Intelligence (AI) for decision support, those experts face the black-box problem and lack trust in AI predictions for fraud. Such an issue has been tackled by employing Explainable AI (XAI) to provide experts with explained AI predictions through various explanation methods. However, fraud detection studies supported by XAI lack a user-centric perspective and discussion on how principles are deployed, both important requirements for experts to choose an appropriate explanation method. On the other hand, recent research in Information Systems (IS) and Human-Computer Interaction highlights the need for understanding user requirements to develop tailored design principles for decision support systems. In this research, we adopt a design science research methodology and IS theoretical lens to develop and evaluate design principles, which align fraud expert's tasks with explanation methods for Explainable AI decision support. We evaluate the utility of these principles using an information quality framework to interview experts in banking fraud, plus a simulation. The results show that the principles are an useful tool for designing decision support systems for fraud detection with embedded user-centric Explainable AI.

show abstract

Models and Tools of Knowledge Acquisition

Joseph

Arun

2021

Computational Management

View full text Add to dashboard Cite

Scenario-Based Requirements Elicitation for User-Centric Explainable AI

Cited by 30 publications

References 89 publications

Deep Learning and Explainable Artificial Intelligence Techniques Applied for Detecting Money Laundering–A Critical Review

Deep Learning and Explainable Artificial Intelligence Techniques Applied for Detecting Money Laundering–A Critical Review

Towards Design Principles for User-Centric Explainable AI in Fraud Detection

Models and Tools of Knowledge Acquisition

Contact Info

Product

Resources

About