Context: Anomaly detection is crucial for maintaining cloud-based software systems, as it enables early identification and resolution of unexpected failures. Given rapid and significant advances in the anomaly detection domain and the complexity of its industrial implementation, an overview of techniques that utilize real-world operational data is needed. Aim: This study aims to complement existing research with an extensive catalog of the techniques and monitoring data used for detecting anomalies affecting the performance or reliability of cloud-based software systems that have been developed and/or evaluated in a real-world context. Method: We perform a systematic mapping study to examine the literature on anomaly detection in cloud-based systems, particularly focusing on the usage of real-world monitoring data, with the aim of identifying key data categories, tools, data preprocessing, and anomaly detection techniques. Results: Based on a review of 104 papers, we categorize monitoring data by structure, types, and origins and the tools used for data collection and processing. We offer a comprehensive overview of data preprocessing and anomaly detection techniques mapped to different data categories. Our findings highlight practical challenges and considerations in applying these techniques in real-world cloud environments. Conclusion: The findings help practitioners and researchers identify relevant data categories and select appropriate data preprocessing and anomaly detection techniques for their specific operational environments, which is important for improving the reliability and performance of cloud-based systems.

English has assumed the role of a global business lingua franca (BELF) at the turn of the 21st century, with an ever-increasing number of multinational corporations (MNCs) adopting English as either their official corporate language, or, the working language as a natural byproduct of a company’s linguascape. This paper investigates the use of English in a business context drawing from the BELF paradigm, i.e. it sets out to compare and contrast the frontstage and backstage English in a multinational organization in written (email) communication, as an answer to a call by Kankaanranta et al. (2018), as this specific kind of study within this genre is still underrepresented and under-researched within the Global South setting. The emails used in this study were collected from a small-sized Turkish-Bosnian international company based in B&H with a total of 10 employees. The approach adopted for the analysis of the study is discourse-analytical in its essence, supported by corpus analysis instruments. The analysis showed that the backstage English, primarily used among employees for internal communication, is indeed in most cases characterized by BELF features. Conversely, frontstage English, was shown to be aligned more closely with native English norms due to its role in corporate branding and external communication, although showing some variability as well. It is expected that the results of the study will help in understanding English communication nuances within this particular business context and help businesses foster clearer, more effective interactions across linguistic and cultural boundaries.

This study investigates the influence of specific surface area (SSA) and aluminum hydroxide particle size on sodium aluminate’s purification efficiency in the Bayer process. This research examines how variations in SSA affect the adsorption and incorporation of contaminants such as Cu, Fe, and Zn, as well as the optimal balance between effective purification and excessive Al2O3 loss. Different SSA values and purification durations are analyzed to optimize the purification process and determine conditions that maximize impurity removal while maintaining system stability. Additionally, solid residue characterization using X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy-dispersive spectroscopy (EDS) provides insights into impurity incorporation mechanisms, including isomorphic replacement, surface adsorption, and co-crystallization. This study highlights key process parameters that influence impurity behavior and crystallization dynamics, offering valuable guidance for refining industrial purification strategies and improving aluminum hydroxide quality.

Abstract Background Humans spend up to 90% of their time indoors and are exposed to a significant number of microbes in their homes, which can have important implications for their health. Aim This study focused on analysing the influence of environmental factors on microbiome diversity and abundance in bed dust and linking the exposure to dust bacteria with asthma. Subjects and methods A total of 90 dust samples were collected from homes of asthmatic patients (n = 59) and controls (n = 31) aged 5–18 years. The bacterial fraction of the microbiome was analysed using 16S rRNA gene high-throughput sequencing on the Illumina MiSeq platform and downstream analyses in QIIME2 and R. Microbiome profiles were associated with asthma and relevant environmental and household data. Results Higher bacterial β-diversity in the environment was shown to be inversely associated with asthma (p = 0.009). Also, living environment (p = 0.002), housing type (p = 0.004), presence of pets in the household (p = 0.001), and cleaning practices (p = 0.006 for dusting and p = 0.011 for vacuuming) were prominent environmental factors affecting the bed dust microbiome. Conclusion Our results suggest significant differences in bacterial community composition between individuals with and without asthma and the interaction between indoor microbiome and asthma is mediated by environmental factors in the household.

Accurate channel estimation is critical for high-performance Orthogonal Frequency-Division Multiplexing systems such as 5G New Radio, particularly under low signal-to-noise ratio and stringent latency constraints. This letter presents HELENA, a compact deep learning model that combines a lightweight convolutional backbone with two efficient attention mechanisms: patch-wise multi-head self-attention for capturing global dependencies and a squeeze-and-excitation block for local feature refinement. Compared to CEViT, a state-of-the-art vision transformer-based estimator, HELENA reduces inference time by 45.0\% (0.175\,ms vs.\ 0.318\,ms), achieves comparable accuracy ($-16.78$\,dB vs.\ $-17.30$\,dB), and requires $8\times$ fewer parameters (0.11M vs.\ 0.88M), demonstrating its suitability for low-latency, real-time deployment.

The development of smart grids poses great challenges to the scientific and professional community. Increasingly strict requirements from regulators and consumers require appropriate actions from the Distribution System Operator (DSO), infrastructure development, and large investments in the modernization and digitalization of electrical distribution systems. The connection of a large number of electricity sources to the existing distribution grid causes problems that are reflected in unauthorized voltage changes or overloads in the network, as well as compromised power quality. Communication infrastructure, as well as the technologies themselves, are often not satisfactory for the requirements that arise in real networks, and the development of smart grids requires appropriate/advanced information and communication infrastructure. The development of smart grids requires an interdisciplinary approach, experts of different profiles, and clearly defined long-term strategies. This paper provides an overview of existing AI technologies which are proposed for application in power systems, as well as an overview of areas where AI can be implemented to support the operation of power systems in the future (such as maintenance, forecasting, optimization, protection, etc.). In a separate section, a simulation of the production of small PV systems connected to consumer households in weak low-voltage grids (LVG) is presented as an illustrative example. An overview of proposed AI applications in LVGs is provided, along with a discussion of possible improvements and overcoming issues that arise in existing grids with prosumers.

The rapid growth of clinical explainable AI (XAI) models raised concerns over unclear purposes and false hope regarding explanations. Currently, no standardised metrics exist for XAI evaluation. We developed a clinician-informed, 14-item checklist including clinical, machine and decision attributes. This is the first step toward XAI standardisation and transparent reporting XAI methods to enhance trust, reduce risks, foster AI adoption, and improve decisions to determine the true clinical potential of applied XAI.

Introductory programming courses are widely known for their difficulty among students. Success in courses is commonly measured in the form of final grades, which might not capture the challenges students face during their learning process. In this paper, we predict students’ success and their future compiler errors based on previously made errors. Furthermore, we examine the effect of applying two clustering techniques before making the predictions and identify key weeks and errors that have the greatest impact on predictions. Experimental results show that students’ compiler errors observed through the semester are an important predictor of students’ achievement and future struggles. Predictions are further improved using sentence encoder-generated embeddings with K-Means algorithm. Our study suggests that students’ errors, particularly the most recent ones, enable meaningful clustering that enhances performance prediction after only three weeks of the semester.

Introductory programming courses present significant challenges for novice learners, often leading to frustration and difficulty in identifying learning gaps. This research aims to develop an AI-driven tool that provides personalized guidance, moving beyond traditional "one-size-fits-all" approaches. Recognizing the limitations of relying solely on digital interaction logs in the era of generative AI, we explore the integration of student personal characteristics and fine-grained programming interactions to predict learning behavior and performance. We will investigate how to accurately predict student outcomes early in the semester, analyze the dynamics of learning behaviors, and design an AI-assisted tool to recommend tailored learning materials and feedback. Our goal is to foster effective learning and mitigate the risks associated with over-reliance on general-purpose AI, ultimately enhancing knowledge retention and problem-solving skills.

Introduction: Tumor-infiltrating lymphocytes (TIL) are linked to responses to chemotherapy and immunotherapy and clinical outcomes, especially in high-risk breast carcinomas. MammaPrint® (MP) and BluePrint® (BP) are genomic tests designed to provide risk stratification and molecular classification for early-stage hormone receptor (HR)-positive breast carcinomas, which could include tumors with HER2-low expression. We investigated correlations between TIL measurements, HER2 status, and MP/BP assays in early-stage HR-positive breast carcinomas. Materials and Methods: 167 early-stage HR-positive breast carcinomas with known MP/BP risk categorization were evaluated for TIL using whole slide scanned images according to the International TILs Working Group 2014 guidelines. HER2-low breast cancers were identified by IHC scores of 1+ and 2+ without HER2 amplification. A subset of high-TIL, high-risk cases underwent TSO500 (Illumina) next-generation sequencing (NGS). Results: The patients had a mean age of 51 years, ranging from 26 to 75 years. Among the profiled cases, 97% were either luminal A (96/167) or luminal B (66/167) breast carcinomas, with only five cases classified as HER2-enriched (n = 2) or basal-like (n = 3) carcinomas. Tumor grade was strongly associated with recurrence risk (p<0.001). The prevalence of the HER2-low phenotype was 65%, including 46/69 (67%) high-risk cases. TIL levels ranged from 0 to 70% and were low (≤10%) in the majority (75%) of cases in the cohort. However, high TIL levels were more frequently observed in cases with high recurrence risk (56% vs. 39%, p = 0.03). Additionally, TIL-enriched high-recurrence risk carcinomas contained targetable genomic alterations, including PIK3CA, BRCA1, BRCA2, and HER2 mutations. Conclusions: TIL levels are higher in early-stage HR-positive breast carcinomas with a high recurrence risk. These tumors also harbor targetable genomic alterations, suggesting that TIL measurement and genomic profiling could enhance risk stratification and identify patients who might benefit from targeted therapies. Her-2 low expression in high-risk patients provides a consideration for including novel ADC therapies in this subset of patients. Citation Format: Zoran Gatalica, Inga Rose, Faruk Skenderi, Nataliya Kuzmova, Semir Beslija, Timur Ceric, Inga Marijanovic, Ilir Kurtishi, Semir Vranic. High Tumor-Infiltrating Lymphocyte Levels Correlate with High MammaPrint® Recurrence Risk in Early-Stage Breast Carcinomas [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P1-11-17.