Contributions in Merlin

Recently, a growing number of credibility assessment technologies (CATs) have been developed to assist human decision-making processes in evidence-based investigations, such as criminal investigations, financial fraud detection, and insurance claim verification. Despite the widespread adoption of CATs, it remains unclear how CAT and human biases interact during the evidence-collection procedure and affect the fairness of investigation outcomes. To address this gap, we develop a Bayesian framework to model CAT adoption and the iterative collection and interpretation of evidence in investigations. Based on the Bayesian framework, we further conduct simulations to examine how CATs affect investigation fairness with various configurations of evidence effectiveness, CAT effectiveness, human biases, technological biases, and decision stakes. We find that when investigators are unconscious of their own biases, CAT adoption generally increases the fairness of investigation outcomes if the CAT is more effective than evidence and less biased than the investigators. However, the CATs' positive influence on fairness diminishes as humans become aware of their own biases. Our results show that CATs' impact on decision fairness highly depends on various technological, human, and contextual factors. We further discuss the implications for CAT development, evaluation, and adoption based on our findings.

Despite the growing interest in explainable AI in the RecSys community, the evaluation of explanations is still an open research topic. Typically, explanations are evaluated using offline metrics, with a case study, or through a user study. In my research, I will have a closer look at the evaluation of the effects of explanations on users. I investigate two possible factors that can impact the effects reported in recent publications, namely the explanation design and content as well as the users themselves. I further address the problem of determining promising explanations for an application scenario from a seemingly endless pool of options. Lastly, I propose a user study to close some of the research gaps established in the surveys and investigate how recommender systems explanations impact the understanding of users with different backgrounds.

Ransomware has remained one of the most notorious threats in the cybersecurity field, for which Moving Target Defense (MTD) has been proposed as a novel defense paradigm. Although various approaches leverage MTD, few of them rely on the operating system and, specifically, the file system, thereby making them dependent on other computing devices, rendering defense against certain threats unrealistic. File-based approaches are less studied here while showing limitations in resource usage and defense effectiveness. Furthermore, existing ransomware defenses merely restore data or detect attacks without preventing them. Thus, this paper introduces the MTFS file system and the design and implementation of three novel MTD techniques – one delaying attackers, one trapping recursive directory traversal, and another one hiding file types. The effectiveness of the techniques is shown in three experiments. First, it is demonstrated that the techniques can delay and mitigate ransomware on real IoT devices. Secondly, in a broader scope, the solution was confronted with 13 ransomware samples, highlighting that it can save 97% of the files. Regarding overhead, the defense system consumes only a small amount of resources, highlighting the feasibility of proactive defense.

We investigate whether banks’ initial responses during the first wave of the COVID-19 pandemic in supporting their customers, communities, and governments were perceived as value-enhancing by investors. Using a unique responsible banking measure for a sample of the largest US and European commercial banks, we find a negative relationship between responsible bank behavior and stock market performance, particularly in the first wave of the pandemic. We also find that riskier banks were affected more negatively if they behaved responsibly. Overall, our findings show that banks’ responsible behavior during a crisis reduces, or at best is not relevant to, shareholder value.

The effectiveness of Voting Advice Applications (VAA) is often compromised by the length of their questionnaires. To address user fatigue and incomplete responses, some applications (such as the Swiss Smartvote) offer a condensed version of their questionnaire. However, these condensed versions can not ensure the accuracy of recommended parties or candidates, which we show to remain below 40%. To tackle these limitations, this work introduces an adaptive questionnaire approach that selects subsequent questions based on users’ previous answers, aiming to enhance recommendation accuracy while reducing the number of questions posed to the voters. Our method uses an encoder and decoder module to predict missing values at any completion stage, leveraging a two-dimensional latent space reflective of political science’s traditional methods for visualizing political orientations. Additionally, a selector module is proposed to determine the most informative subsequent question based on the voter’s current position in the latent space and the remaining unanswered questions. We validated our approach using the Smartvote dataset from the Swiss Federal elections in 2019, testing various spatial models and selection methods to optimize the system’s predictive accuracy. Our findings indicate that employing the IDEAL model both as encoder and decoder, combined with a PosteriorRMSE method for question selection, significantly improves the accuracy of recommendations, achieving 74% accuracy after asking the same number of questions as in the condensed version.

Threat modeling has been successfully applied to model technical threats within information systems. However, a lack of methods focusing on non-technical assets and their representation can be observed in theory and practice. Following the voices of industry practitioners, this paper explored how to model insider threats based on business process models. Hence, this study developed a novel insider threat knowledge base and a threat modeling application that leverages Business Process Modeling and Notation (BPMN). Finally, to understand how well the theoretic knowledge and its prototype translate into practice, the study conducted a real-world case study of an IT provider's business process and an experimental deployment for a real voting process. The results indicate that even without annotation, BPMN diagrams can be leveraged to automatically identify insider threats in an organization.

How might irrelevant events infiltrate voting decisions? The current research introduces a new mechanism - regulatory focus—by which incidental environmental factors can affect vote choice. Regulatory focus theory proposes that there are two fundamental psychological orientations in how people navigate their worlds: A prevention focus tunes cognition towards security, safety, protection, and risk aversion, whereas a promotion focus orients attention toward achieving growth and positive outcomes. We present a model for how wind speed on Election Day affects voting by shifting the regulatory focus of voters. We propose that increased wind speed shifts voters toward selecting prevention-focused options (e.g., restricting immigration, rejecting Brexit, rejecting Scottish Independence) over promotion-focused options (e.g., promoting immigration, favoring Brexit, favoring Scottish Independence). We further argue that wind speed only affects voting when an election clearly offers a choice between prevention and promotion-focused options. Using a mixed-method approach—archival analyses of the “Brexit” vote, the Scotland independence referendum, and 10 years of Swiss referendums, as well as one field study and one experiment - we find that individuals exposed to higher wind speeds become more prevention-focused and more likely to support prevention-focused electoral options. The findings highlight the political importance of incidental environmental factors. Practically, they speak to the benefit of absentee voting and expanding voting periods beyond traditional election days.

Many blockchain consortia have been established to build blockchain information systems. While the developed blockchain information systems were promising, few have reached market entry. Indeed, blockchain consortia often lost development focus due to high system complexity and a lack of understanding of how to create a system that will serve the needs and bring value to all stakeholders. Thus, stakeholders struggled to leverage blockchain information systems' full value. Prior studies demonstrated that blockchain systems pose not only technical but also organizational challenges. Analysing six blockchain consortia, we identify their value mechanisms, organizational problems, and organizational solutions that successful blockchain consortia experience while organizing themselves for value. As a result, we propose a new organizational form, i.e., a layered organization, for blockchain consortia to achieve better value creation.

Climate change will impact current and future generations in different regions very differently. This paper develops the first large-scale, annually calibrated, multi-region, overlapping generations model of climate change and carbon policy. It features region-specific temperature and damage functions with the phased impact of emissions on global and regional temperature calibrated to the latest scientific evidence. Absent policy, climate change may, under high-damage scenarios, dramatically reduce GDP in most regions, with India, Brazil, and the South Asian Pacific suffering long-term catastrophic damages. Carbon taxation, coupled with region- and generation-specific transfers, can both correct the carbon externality and raise the welfare of all current and future agents across all regions by 4.3 percent. The impact on the use and duration of fossil fuels is dramatic as is the reduction in the path of global emissions. However, achieving completely uniform welfare gains leaves future generations in particular regions facing exceptionally high compensatory payments. Fortunately, a carbon tax-cum redistribution policy that limits this burden for any generation in any region to less than 10 percent, measured on a consumption-equivalent basis, can deliver a 4.0 percent or higher welfare gain for all peoplekind — present and future. However, carbon taxes set through time, at carbon’s marginal social cost, do far too little to mitigate climate change unless all major emitters, particularly China, adopt them and do so immediately.

This study examines the alpha-generating power of the public-report-based ESG score, which is based on ESG incident data collected by RepRisk from various public sources, and its relationship with the self-disclosure-based ESG score obtained from Refinitiv. The authors construct pure ESG factor portfolios to neutralize exposure to common style factors and isolate the pure ESG factor returns. Their results suggest that (i) the source difference is the main cause of the negative correlation between the public report ESG and the self-disclosure ESG score, (ii) the public report ESG score and its subscores produce the mixed results in terms of their adjusted factor returns across regions, and (iii) the combination of the public report ESG and the self-disclosure ESG score significantly improves the risk–return profiles of the combined ESG factor returns in the United States, European Union (EU), and Japan.