Accurate extraction of functional areas in port waters is essential for enhancing port operational oversight, optimizing vessel scheduling, and supporting maritime safety. However, existing approaches often rely on supervised learning, extensive parameter tuning, and labeled datasets, limiting their scalability, adaptability, and operational efficiency. To address these gaps, this study proposes PortMiner, a novel unsupervised data mining framework that systematically extracts functional areas from raw vessel trajectory data without requiring manual annotations. The framework introduces a Spatio-Temporal Adaptive Sliding Windows (STASW) method to detect stop behaviors dynamically, using self-adaptive parameters derived directly from the data. Trajectories are first encoded into geohash-based sequential grids, enabling efficient detection of stop and port inbound/outbound behaviors. Functional zones such as berths, anchorages, and navigational channels are then delineated through multi-level spatial aggregation and connectivity-based clustering. Experimental results on benchmark datasets show that STASW achieves 98.83% accuracy, outperforming state-of-the-art deep learning methods, while significantly reducing computational time and cost. Validation against official nautical charts confirms PortMiner’s high fidelity in identifying port-functional structures. The extracted results are also made publicly accessible via an interactive platform ( https://portminer.netlify.app/ ), offering practical insights for intelligent port operation and maritime logistics planning.

Frequent itemset mining is a core data mining task aimed at uncovering recurrent patterns within transactional databases. Traditional methods rely on a minimum support threshold, which is often difficult to determine. Top- k mining offers a pragmatic alternative by retrieving the k most frequent itemsets. We propose two novel algorithms: HTK-Miner and HTK-negFIN. HTK-Miner, based on equivalence class theory and breadth-first search, utilizes vertical structures with four operational modes (TS, BSN, DTS, and DBSN). Its key innovation, the Quick Heap (Q-Heap), dynamically raises the support threshold to enable early pruning and accelerated identification. Furthermore, HTK-Miner requires a single database scan and employs compressed representations to reduce execution time and memory usage. HTK-negFIN adapts the pattern-growth paradigm by extending the efficient negFIN algorithm to the Top- k framework, integrating the Q-Heap and shared optimizations to achieve high performance. Experiments on diverse benchmark datasets demonstrate that our proposed algorithms consistently outperform state-of-the-art methods in both runtime and memory efficiency. These results highlight HTK-Miner and HTK-negFIN as scalable, effective solutions for Top- k frequent itemset mining.

A new classification system of beer categories and styles based on large-scale data mining and self-organizing maps of beer recipes

The functional performances are encoded by protein structures, and modified structure-based strategies for customizing food proteins have major implications for the food industry. The glycation reaction that typically occurs between food components is a promising strategy for protein modification due to its mild reaction conditions and natural occurrence during processing. However, the complexity and dynamic nature of glycation reactions hinder precise control, and there is a large imbalance between abundant structural data and function information. Artificial intelligence (AI), with its capacity for large-scale data integration and predictive modeling, offers transformative potential for elucidating glycation-structure-function relationships. This review therefore aims to (1) summarize advances in analytical strategies for glycated proteins, highlighting techniques for site localization, conformational analysis, and multi-source data mining; (2) elucidate how glycation-induced structural modifications alter protein functional performance, providing mechanistic insights into physicochemical properties and biological activities; and (3) discuss emerging AI-driven approaches, including deep learning and inverse design, for predicting and optimizing glycation patterns. These insights provide a systematic framework to accelerate rational development of functional proteins and promote innovative applications in the food industry.

Impact of pharmacist-led digital interventions on asthma outcomes: A systematic review.

Realistic-NPs trigger depression-like behaviors via mitochondrial iron overload mediating ferroptosis.

A novel multisource-intelligent calibration method for discrete element model parameters and application in macro-mesoscopic strength analysis of slope soil

Nanoparticles have gained substantial attention in view of their distinctive physicochemical attributes and widespread applications in several fields. However, the prompt development and extensive consumption of nanotechnology may provoke inexorable diffusion of nanoparticles into the environment, associated with potential toxic effects. Hence, the preliminary toxicological screening of nanomaterials becomes indispensable for their harmless utilization and ecological safety. Nanotoxicology deals with the study of undesirable effects attributed to nanoparticles. It includes the nature, intensity, and characteristics of toxic insult caused by individual or combined use of nanoparticles. Nanoinformatics represents a systematic approach for collecting, organizing, validating, storing, sharing, visualizing, modeling, and analyzing data from nanotechnology processes and materials. The conventional nanotoxicity assessment methods using invitro assays or animal models are time-consuming and relatively expensive, whereas computational modeling of physicochemical properties and existing toxicity data can be effectively used to determine the safety of nanomaterials. Nanoinformatics involves the integration of nanospecific databases (e.g., NanoDatabank, eNanoMapper, Data and Knowledge on Nanomaterials, Online Chemical Modeling Environment, and Nanoparticle Information Library) with modeling frameworks such as quantitative nanostructure-activity/toxicity relationship, molecular docking, physiologically based toxicokinetic models, and molecular dynamics simulation for predictive nanotoxicity assessment. Moreover, the process of computer-aided nanotoxicity prediction can be further expedited using the latest data mining techniques. Challenges in collecting sufficient, high-quality, nanotoxicity data, as well as in standardizing the training data sets require careful consideration to further expand the applications of nanoinformatics techniques in predictive nanotoxicology. This article highlights the current status and future perspective of nanoinformatics-based predictive toxicological screening of nanomaterials.

Accounting fraud is a big problem that affects companies, investors, and the overall economy, particularly in Indian listed companies where money transactions are now more complex and done by online. If companies give wrong financial information, report incorrect income, and hide their liabilities it misleads stakeholders and reduce trust in financial systems. This research highlights on how artificial intelligence (AI) can make fraud detection in accounting faster and with better accuracy. Conventional fraud detection methods mainly depend on manual review and simple auditing techniques, which take a lot of time and may not to be determine hidden or sophisticated frauds. Compared to older methods, AI tools like machine learning, data mining, and predictive analytics can quickly study large amount of financial data and spot unusual patterns, trends, and irregularities that may suggest fraud. This system can also learn from previous information and become more accuracy over time. The study examines on company data from selected Indian listed companies to examine the effectiveness of AI-based models can detect fraud financial statements. It also shows the role of AI in helping auditors, improving risk management, and making company control systems. The findings highlights that AI increases the speed and accuracy of fraud detection but also helps in identify fraud early and prevent it. In overall study concludes that the benefits of artificial intelligence in accounting can strongly reduce the occurrence of fraud and improve company financial report quality, and strengthen trust among investors and stakeholders.

Background. Many pharmaceuticals, including antibiotics, diuretics, some antitumor agents, hormones, etc., can promote the depletion of magnesium (Mg), pyridoxine (vitamin B6, VB6), and other micronutrients (MNs) in the body. This process may lead to the development of hypomagnesemia and concomitant MN deficiencies, which are associated with a range of adverse effects, including neurotoxicity, cardiotoxicity, hepatotoxicity, etc. Moreover, the resulting micronutrient deficiency (MND) may paradoxically aggravate the underlying pathophysiological mechanisms of the diseases for which these drugs are prescribed, thereby potentially diminishing therapeutic efficacy and contributing to treatment-related complication. Objective: Chemoreactomic assessment of anti-micronutrient (anti-MN) effects of all drugs included in the Anatomical Therapeutic Chemical (ATC) classification system. Material and methods. Using modern data mining techniques, including mathematical approaches from topological data analysis, labeled graph theory (chemographs), and related method, this study performed a systematic computer-based analysis of databases describing the Mg-depleting effects of drugs; original algorithms for numerically predicting the Mg- and VB6-removing effects of drugs. Original algorithms were developed for the numerical prediction of Mg- and VB6-depleting properties of drugs, as well as for the assessment of other anti-MN effects. These algorithms were subsequently applied in a chemoreactomic screening of 2,527 drugs classified within the ATC system. Results. A database describing anti-MN properties of drugs was created for 24 MN balance indicators for 18 MNs. Algorithms for predicting the anti-MN properties of drugs were developed with a classification accuracy of 92±10% in cross-validation (the accuracy of predicting VB6 MND – 88%, Mg MND – 94-98%). On average, each drug from the ATC group accounts for 8.5±6.5 anti-MN effects. Only 100 out of 2527 (4%) drugs did not exhibit a negative impact on MN, primarily amino acids, MNs themselves, and choline drugs. The most pronounced negative impact of the drugs under study was related to the metabolism of vitamin D3 (505 ATC categories), VB6 (475 ATC categories), iron (419 ATC categories), vitamin B1 (386 ATC categories), and Mg (375 ATC categories). VB6 MND was caused by 1701 drugs, Mg MND – by 1064 drugs. Antibiotics for systemic use (ATC code J01), psycholeptics (N05) and psychoanaleptics (N06), antineoplastic agents (L01), sex hormones and modulators of the reproductive system (G03), analgesics (N02), antidepressants (N06A), diuretics (C03), antihistamines for systemic use (R06A), anti-inflammatory and antirheumatic agents (M01), direct-acting antivirals (J05A), and antiepileptic agents (N03A) were found to affect adversely the homeostasis of both Mg and VB6. A detailed description of the anti-Mg and anti-VB6 properties of these drug classes was provided. The data obtained via chemoreactomic analysis were compared with that obtained by experimental and clinical studies of Mg and VB6 preparations. Conclusion. The conducted chemoreactomic analysis provides a substantiated basis for supporting pharmacotherapy with selected medicinal preparations based on organic salts of Mg and VB6.

ABSTRACT Feature selection poses a common challenge in data mining and machine learning. This paper offers a comprehensive overview of the field, encompassing its historical context, recent advancements, and associated challenges. Topics covered include categorization of feature selection methods, development of ensemble frameworks, and assessment of selection stability. Current research trends in feature selection and computational intelligence are examined, alongside practical applications in ensemble configuration, threshold detection, evaluation criteria, and stability enhancement techniques. This article is categorized under: Algorithmic Development > Ensemble Methods

Identification and characterization of a novel lysin-like endopeptidase from the Vibrio cholerae predator phage ICP1.

Aim: This study aims to investigate how Human Resource (HR) analytics can enhance key stages of the employee lifecycle within large public sector organizations. The focus is recruitment, performance management, engagement, and retention. It is situated within the evolving domain of public workforce management, where the integration of HR analytics technologies has become increasingly important, yet a gap persists in comprehensive and integrated approaches capable of managing the entire employee lifecycle effectively. Methods: To address this gap, the study employs a mixed-methods approach, utilizing advanced data mining techniques, rigorous statistical analyses, and validation protocols applied to extensive HR datasets drawn from multiple large-scale public workforce systems. Results: The findings indicate that the application of HR analytics significantly enhances decision-making capabilities, resulting in improved employee lifecycle outcomes, including optimized talent acquisition, increased employee engagement, reduced turnover rates, and overall cost efficiencies. Conclusion: These results highlight the transformative potential of HR analytics in public sector human resource management, demonstrating its capacity to improve organizational effectiveness and workforce outcomes when applied in an integrated and systematic manner. Recommendation: The study recommends the development of policy frameworks that support robust data governance and ethical considerations, alongside continued innovation in HR practices to fully leverage analytics capabilities. It further highlights the need for future research through longitudinal studies and the exploration of AI-driven predictive models to sustain workforce excellence.

Background: The transition from fossil-derived polymer additives to renewable alternatives is essential to mitigate environmental persistence and ensure chemical safety within the plastics industry. This review provides a comprehensive overview of recent developments in bio-based functional additives and their integration into circular economy frameworks. Methods: Following PRISMA guidelines, a systematic literature search was conducted using the Scopus database for studies published between 2023 and 2026. Search terms targeted bio-based plasticizers, flame retardants, antioxidants, and compatibilizers. Studies were screened against predefined inclusion criteria, specifically focusing on experimental validation in polymer matrices, while data mining was employed to map emerging research fronts. Results: From an initial 996 records, 54 studies were selected after removing duplicates and ineligible articles. The findings highlight a paradigm shift from passive physical fillers toward active, multifunctional macromolecular agents. Recent literature demonstrates that targeted molecular interventions, such as phosphorylated lignin and biomimetic structures, can resolve trade-offs between ductility and thermal stability at low loadings (<5 wt%). Synthesis routes, performance outcomes, and end-of-life trajectories for each additive class are summarized. Conclusions: Bio-based additives have evolved from simple substitutes into strategic tools for the molecular programming of sustainable polymers. Although challenges regarding scalability and high-temperature processing persist, their integration into circular economy strategies establishes a clear roadmap for next-generation bioplastics.

The study examines methods of explainability and semantic verification for financial data mining outcomes in computer decision support systems, particularly in high-risk industries such as aerospace. The purpose of the article is to analyze modern explainability methods and approaches to verifying the results of intelligent systems within a financial context, identify their limitations, and justify an approach to explainable semantic verification based on a combination of xAI methods, ontological knowledge representation, and formal verification procedures. Tasks include: analyzing modern methods of explainability and approaches to verifying the functioning of intelligent systems; identifying the limitations of existing solutions in ensuring the logical admissibility, semantic consistency, and reliability of data mining, results; and developing an approach to explainable semantic verification specifically for financial data. The study employs methods of analyzing and generalizing scientific sources, systemic and comparative analysis, and approaches to ontological modeling and the semantic interpretation of machine learning results. The findings indicate that modern xAI approaches provide interpretations of machine learning outputs, but do not guarantee their logical admissibility, semantic consistency, or regulatory acceptability in the financial sphere. The feasibility of integrating xAI methods with ontological knowledge models and formal verification procedures is substantiated, allowing for expanded quality control of analytical conclusions. The proposed approach involves the sequential implementation of analytical result formation, explanation generation, semantic mapping, logical verification, and result reliability assessment. Conclusions. The scientific novelty of the obtained results lies in the justification of the proposed approach to explainable semantic verification of financial data mining results. Unlike existing methods, this approach provides not only a meaningful interpretation of the output, but also its logical verification and semantic consistency with domain-specific knowledge and industry constraints.

With the exponential growth of sensing technology and computing power, artificial intelligence (AI) is increasingly deeply involved in the field of sports prediction, driving a paradigm shift in sports scientific research from "experience-driven" to "data-driven". This article systematically reviews the evolution logic of AI in sports prediction and analyzes the underlying mechanism of the migration from traditional statistical models to deep learning and reinforcement learning algorithms. The study finds that AI, through the deep mining of multi-source heterogeneous sports data, not only demonstrates remarkable accuracy in predicting competitive performance and assessing the risk of sports injuries, but also achieves functional reconfiguration in the dimensions of sports industry decision support and enhanced spectator experience. This article constructs a technical path model for AI sports prediction and conducts a deep examination of its value spillover and technical limitations in practical applications. The research aims to provide theoretical support for the digital transformation of China's sports industry and offer path references for the optimization and ethical governance of intelligent sports prediction algorithms.

Research on Production Data Mining and Steam Injection Timing Intelligent Decision System for Heavy Oil Thermal Recovery Well Groups

Predicting student academic performance has become increasingly vital in the field of educational data mining, as institutions seek data-driven strategies to enhance learning outcomes. However, many existing models rely solely on behavioral indicators or static features, often overlooking the role of time and context in shaping learning behavior. This limitation reduces predictive accuracy and adaptability in academic environments. To address this challenge, this study introduces EduFuseNet, a hybrid deep learning framework that integrates behavioral and spatiotemporal data for accurate classification of student performance. The workflow begins with data collection from a Student Academic Performance dataset, comprising both behavioral metrics and spatiotemporal information. The raw data undergoes preprocessing, including missing value imputation, one-hot encoding of categorical variables, and min-max scaling of numerical features. The processed data is then passed through two specialized branches: a Tabular Neural Structure-Aware (TabNSA) module that captures complex interdependencies within behavioral data, and a Spatiotemporal Transformer module that models temporal and sequential patterns in learning activities. The feature embeddings from both branches are fused and passed through fully connected layers to generate predictions across five academic performance bands, enabling precise classification and early risk identification. EduFuseNet achieved an accuracy of 99.00%, with a precision of 99.04%, recall of 99.00%, and F1-score of 99.01%, reflecting strong and reliable predictive performance. By leveraging both behavioral and temporal learning indicators, the model serves as an effective tool for early academic monitoring and intervention.

Retraction Note: a literature survey on various aspect of class imbalance problem in data mining

Pattern discovery in early monsoon rainfall across Kerala using data mining techniques