This study critically examines the methodological rigor in credit card fraud detection research, revealing how fundamental evaluation flaws can overshadow algorithmic sophistication. Through deliberate experimentation with improper evaluation protocols, we demonstrate that even simple models can achieve deceptively impressive results when basic methodological principles are violated. Our analysis identifies four critical issues plaguing current approaches: (1) pervasive data leakage from improper preprocessing sequences, (2) intentional vagueness in methodological reporting, (3) inadequate temporal validation for transaction data, and (4) metric manipulation through recall optimization at precision's expense. We present a case study showing how a minimal neural network architecture with data leakage outperforms many sophisticated methods reported in literature, achieving 99.9\% recall despite fundamental evaluation flaws. These findings underscore that proper evaluation methodology matters more than model complexity in fraud detection research. The study serves as a cautionary example of how methodological rigor must precede architectural sophistication, with implications for improving research practices across machine learning applications.

Link Prediction (LP) is an essential task over Knowledge Graphs (KGs), traditionally focussed on using and predicting the relations between entities. Textual entity descriptions have already been shown to be valuable, but models that incorporate numerical literals have shown minor improvements on existing benchmark datasets. It is unclear whether a model is actually better in using numerical literals, or better capable of utilizing the graph structure. This raises doubts about the effectiveness of these methods and about the suitability of the existing benchmark datasets. We propose a methodology to evaluate LP models that incorporate numerical literals. We propose i) a new synthetic dataset to better understand how well these models use numerical literals and ii) dataset ablations strategies to investigate potential difficulties with the existing datasets. We identify a prevalent trend: many models underutilize literal information and potentially rely on additional parameters for performance gains. Our investigation highlights the need for more extensive evaluations when releasing new models and datasets.

Recent advancements in artificial intelligence (AI) systems, including large language models like ChatGPT, offer promise and peril for scholarly peer review. On the one hand, AI can enhance efficiency by addressing issues like long publication delays. On the other hand, it brings ethical and social concerns that could compromise the integrity of the peer review process and outcomes. However, human peer review systems are also fraught with related problems, such as biases, abuses, and a lack of transparency, which already diminish credibility. While there is increasing attention to the use of AI in peer review, discussions revolve mainly around plagiarism and authorship in academic journal publishing, ignoring the broader epistemic, social, cultural, and societal epistemic in which peer review is positioned. The legitimacy of AI-driven peer review hinges on the alignment with the scientific ethos, encompassing moral and epistemic norms that define appropriate conduct in the scholarly community. In this regard, there is a "norm-counternorm continuum," where the acceptability of AI in peer review is shaped by institutional logics, ethical practices, and internal regulatory mechanisms. The discussion here emphasizes the need to critically assess the legitimacy of AI-driven peer review, addressing the benefits and downsides relative to the broader epistemic, social, ethical, and regulatory factors that sculpt its implementation and impact.