Search Results (79)

Large language models (LLMs) are increasingly used in programming tasks, including code generation, debugging, validation, optimization, test generation, and educational support. Their use as graders of student programming assignments is, however, still not fully understood, especially in object-oriented programming (OOP). In this context, correctness depends not only on the final output of a program but also on class structure, encapsulation, abstraction, and design decisions. This paper introduces a dataset of 271 C++ student solutions to eight OOP examination problems, together with the problem statements and the reference grades produced by the university auto-grader. We evaluate four locally served open-weight models—phi4-mini, gpt-oss:20b, qwen3.6:27b, and qwq:32b—under one endpoint, one prompt, and deterministic decoding, and we report four properties that a deployable grader must have. The first is output-format compliance, which ranges from 88.9% to 98.9% across models: qwq:32b failed to terminate on 30 of 271 submissions even at a doubled generation budget, spending 7.7 MJ, 29.9% of its total energy, to produce nothing. The second is agreement with the reference grades on the 30-point course scale, where qwen3.6:27b performs best (MAE 3.23 points, $r=0.85$) and phi4-mini worst (MAE 10.90, $r=0.17$); all four models grade systematically more strictly than the auto-grader. The third is reproducibility, which fails at two levels. Repeating an identical run, three models returned identical grades at temperature 0 with a fixed seed while gpt-oss:20b varied by up to 21 points on the same submission; restricting the serving daemon to one concurrent request makes it reproducible, locating that failure in request batching rather than in sampling. Grading the same submissions as part of a different set, however, no model is reproducible: agreement with the original run ranges from 13 to 33 of 40 submissions, and a control run with the original client excludes our instrumentation as the cause. The fourth is degeneracy: phi4-mini awarded an identical grade of zero to every submission in three of the eight examinations, 109 submissions in total, so its aggregate agreement statistics describe a model that is not grading. On a subset graded blind by two instructors, each instructor agreed with the auto-grader more closely than the two agreed with each other, so the residual error of the best-performing model is of the same order as ordinary disagreement between human graders. We also report a negative result: a declaration-level reference could be reconstructed for only five of the eight examinations, because the auto-grader merges missing declarations and wrong return values into single checks. Together these results indicate that agreement with a reference is an insufficient basis for choosing a local grading model, and that compliance, determinism, degeneracy, and the cost of failed generations must be measured alongside it. Full article

Cities increasingly use resilience policies to manage climate risk, but it remains unclear how these policies alter firms’ green strategies. Using a difference-in-differences design, we analyze 2011–2023 panel data for A-share listed firms around China’s Climate-Resilient City Construction (CRCC) pilot. CRCC exposure is associated with a higher text-based measure of corporate green transformation. The estimate is stable in parallel-trends and placebo tests, alternative outcome measures, exclusions of confounding policies and shocks, alternative specifications, double machine learning, pre-policy propensity-score matching, and instrumental-variable estimation. Mechanism estimates are consistent with three channels through which CRCC may affect firm behavior. CRCC strengthens executive green cognition, redirects environmental spending toward prevention, and supports green technological and management innovation. By contrast, the estimate for end-of-pipe investment is small and statistically insignificant. A pooled interaction test finds differences across firm life-cycle stages. Comparisons by city size, exemplary-city status, and supply-chain resilience remain descriptive because their grouping indicators are unavailable for pooled re-estimation. Spatial estimates reveal limited, non-monotonic spillovers. Overall, the results indicate that an urban adaptation policy can influence firm strategy, subject to the text-based outcome and the sample of Chinese listed firms. Full article

Background/Objectives: Children now spend a substantial part of daily life in digital environments, and their digital, media, and information literacies shape their online safety, social-emotional development, and mental health, making these competencies a concern for child public health and preventive paediatric care. This study applies a Synthetic Knowledge Synthesis (SKS) to map how digital, media, and information literacies (hereafter digital literacies) among children have evolved, comparing neurotypical and neurodivergent children. SKS is a semi-automated approach that combines descriptive bibliometrics, keyword co-occurrence mapping, and qualitative content analysis to map an entire research field or topic. Methods: We treat digital literacies as relevant to children’s health, well-being, and safe online participation, and we analyse Scopus-indexed literature from 1996 to 2025 using descriptive bibliometrics, keyword co-occurrence mapping, and qualitative content analysis. Results: The mapping shows strong growth in general digital-literacy scholarship alongside a small but persistent body of work on a “double digital divide” affecting children with autism spectrum disorder (ASD), ADHD, and dyslexia. The two bodies of work differ in emphasis: the general literature foregrounds social integration and critical agency, while research on neurodivergent children foregrounds inclusive pedagogy, implicit learning, and multimodal expression. It also shifts from general digital-safety awareness toward protective mediations for vulnerabilities such as social-emotional decoding differences and impulsivity, which bear on children’s online safety and mental health. Because bibliometric mapping reveals where research is concentrated rather than what works in practice, the synthesis points to evidence gaps rather than proven methods. Conclusions: On this basis, we argue for an integrated public-health and socio-educational framework that complements universal literacy standards with adaptive, assistive safety nets, so that all children can take part in digital life safely and on equal terms. Full article

Objectives: Hospital efficiency provides a systems-level perspective for identifying unrealized capacity to improve care delivery, quality, and patient safety. This study evaluated relative technical efficiency among U.S. short-term acute care hospitals (2018–2024) and evaluated how structural IT operating investments are converted into clinical service volume and quality-related performance. Methods: A longitudinal panel of 8589 hospital-year observations was utilized to estimate technical efficiency with an output-oriented Data Envelopment Analysis (DEA) under variable returns to scale. A secondary Simar–Wilson double-bootstrapped truncated regression (n = 2147 complete casesexamined associations with bias-corrected efficiency, clinical quality, case mix, and operational scale. Results: Mean DEA efficiency was 0.567, with 7.38% (n = 634) operating on the annual efficiency frontier and a mean output expansion potential of 102.49%. Efficient hospitals maintained higher IT operating spending. Stage 2 regression showed that higher Hospital Value-Based Purchasing quality performance was significantly associated with greater inefficiency. Conversely, higher case mix complexity and larger operational scale were associated with higher efficiency. Conclusions: Most hospitals operated below the best-practice frontier, indicating gaps in converting resources into service volume and quality. Because core operational drivers were included in the primary DEA model, observed associations show descriptive structural patterns rather than direct cause-and-effect relationships. Full article

This work deals with questions of enhancing the scalability and security of linear chain Bitcoin by introducing a D-BTC (Domino Bitcoin) protocol, supported by a simply connected two-dimensional structure. The paper seeks to answer the question: can the linear topology of Bitcoin be replaced by a richer geometric structure that simultaneously (i) enlarges the number of valid positions where parallel mining can occur, and (ii) strengthens the asymptotic decay of the double-spend reversal probability? In the D-BTC protocol, the blocks, called B-dominoes (Bitcoin dominoes) are organized as a finite connected region subset of ${\mathbb{Z}}^2$ without holes, also called a lattice. Simple connectivity plays a central role in D-BTC and to mine a (valid) B-domino, a miner has to compute four PoW (Proof of Work), corresponding to cardinal directions, allowing them to add it to the frontier of the lattice, under the constraint that the new lattice is simply connected. We introduce a new deterministic consensus based on maximization of the lattice surface. By using a simple version of the isoperimetric inequality, we see that the frontier size grows as $\Omega \left(\sqrt{n}\right)$, where n is the lattice size. Following the Nakamoto’s heuristic, and under the honest majority assumption, a double-spending attack is successful with probability decaying exponentially in $k^2$, where k is the minimum Manhattan distance of the concerned B-domino from the lattice frontier. Additionally, we set up implementations and experiments to demonstrate the practical viability of the protocol with authentic gossip-based message propagation and complete Merkle tree verification. Full article

With the widespread adoption of cryptocurrencies, the ability to conduct continuous offline payments has increasingly become a critical technological requirement. In network-constrained scenarios, current dual-offline payment technologies are useful for single transactions. However, their limitations in continuous payment scenarios have become increasingly evident, making them unable to meet real-world application needs. This has prompted the industry to demand more urgent innovations in research on continuous offline payment capabilities. To address these challenges, this paper proposes a continuous dual-offline payment system capable of supporting multiple continuous payments. The system integrates elliptic curve cryptography (ECC) and zero-knowledge proof (ZKP) technology to generate secure asset credentials, ensuring both immutability and privacy credentials throughout the offline payment lifecycle. A dynamic credential decomposition mechanism enables the splitting of input credentials into change credentials and receipt credentials, facilitating uninterrupted dual-offline payments between hardware wallets. Additionally, it incorporates a batch verification scheme based on smart contracts, utilizing zero-balance verification and chained hash tracing to ensure payment uniqueness and prevent double-spending attacks, thereby guaranteeing the verifiability and validity of payment settlements. Experimental evaluations demonstrate that the proposed system reduces gas consumption per payment and improves execution efficiency during batch processing, combining high security with strong performance. This research provides a feasible solution for the application of digital currencies in offline scenarios, carrying significant theoretical value and practical significance for driving technological innovation and application expansion in the cryptocurrency field. In addition to cryptocurrency payments, the proposed system is also applicable to IoT and sensor network environments. Many IoT devices operate in disconnected or network-limited areas and require secure micro-transactions. Our dual-offline payment mechanism supports such scenarios, as the main cryptographic operations are lightweight enough for typical IoT hardware. This further extends the practical value of our system beyond traditional cryptocurrency payments. Full article

This paper studies the probability of a double-spend attack in an Ouroboros-like Proof-of-Stake (PoS) setting when confirmation decisions must be made for a finite number of blocks. Existing security analyses of Ouroboros-family protocols are mainly asymptotic and therefore do not directly provide the attack probability for a fixed confirmation depth. We consider an analytically tractable model that allows empty slots and multiple slot leaders, and assumes fixed stake distribution within an epoch, one-block growth of the public longest chain in any slot containing at least one honest leader, and next-slot block visibility. These assumptions hold when the time slot length is much greater than the network delay, and are applicable to practical deployment scenarios such as Cardano. Under these assumptions, for the first time, an exact closed-form solution for the success probability of a double-spend attack considering a realistic model with multiple leaders and empty time slots. Numerical examples illustrate how the required confirmation depth depends on the adversarial stake ratio and the active slot coefficient. The results apply to the stated analytical model and do not yet cover delayed fork resolution or the full protocol-level fork-choice and finality mechanisms of Ouroboros Praos. Full article

Social determinants of health (SDoH) are non-medical factors shaped by the socioeconomic status of individuals or communities that influence the onset and progression of diseases and affect their outcomes. We have narratively analyzed the most important findings relating chronic kidney disease (CKD) and SDoH, evaluating the following items: (i) medical care and social determinants of health, (ii) socioeconomic risk for kidney disease at the individual level and (iii) socioeconomic risk for kidney disease at the population level. SDoH can be categorized by how they influence a person’s daily life. Individual factors include personal lifestyle choices such as smoking habits, alcohol consumption, and how a patient spends their non-working time. Community factors include structural elements such as average household income, educational attainment, employment rates, and the quality of the surrounding physical environment. Research consistently shows that a low socioeconomic status is a primary driver of poor clinical outcomes. While healthcare systems vary globally, the negative impact of socioeconomic deprivation on CKD patients remains a constant. Disadvantaged patients experience a faster loss of renal function, and there is a significantly higher incidence of cardiovascular events and mortality compared to those with financial stability. Financial hardship often leads to a “double burden,” where the struggle to afford care triggers a decline in both physical health and mental well-being. To improve patient care, it is essential to raise awareness among healthcare providers regarding the profound impact of these social factors. More precise data and thorough research are needed to fully understand these associations and develop targeted interventions. Full article

The transition from centralized power systems to decentralized infrastructures with a high share of renewable energy sources calls for reliable settlement in P2P electricity trading across “smart” regions. Blockchain platforms can enhance transparency and facilitate automated settlement; however, double-spend attacks still pose a threat to transaction finality and, consequently, undermine trust in the payment layer. This paper quantifies this risk through a probabilistic analysis of classical double-spend scenarios for Proof-of-Work (PoW) and Proof-of-Stake (PoS) blockchains augmented with periodic checkpoints, which render the chain history prior to the latest checkpoint effectively irreversible. We develop attack models for both consensus mechanisms and derive explicit formulas for the attacker’s success probability as a function of the adversarial share, the spacing between checkpoints, and the number of confirmation blocks. On this basis, we compute the minimum confirmation depth needed to satisfy a predefined risk threshold. Numerical evaluation using the derived expressions shows that checkpoints consistently reduce double-spend probability relative to checkpoint-free baselines; in the evaluated settings, the reduction reaches up to 44% and becomes more pronounced as the adversarial share increases. Finally, the analysis yields practical guidance for energy trading applications: accept a payment after the computed number of confirmations when it fits within a single checkpoint interval; otherwise, treat finality as reaching the next checkpoint. Full article

The entire system collapses due to the issues of inadequate centralized storage capacity, poor scalability, low storage efficiency, and susceptibility to single point of failure brought on by huge power consumption data in the smart grid; thus, an alliance chain-driven multi-cloud storage and verifiable deletion method for smart grid data is proposed. By leveraging the synergy between alliance blockchain and multi-cloud architecture, the encrypted power data originating from edge nodes is dispersed across a decentralized multi-cloud infrastructure, which effectively mitigates the danger of data loss resulting from single-point failures or malicious intrusions. The removal of expired and user-defined data is guaranteed through a transaction deletion algorithm integrated into the indexed storage deletion chain and strengthens the flexibility and security of the storage architecture. Based on the Practical Byzantine Fault-Tolerant Consensus Protocol with Ultra-Low Storage Overhead (ULS-PBFT), by the hierarchical grouping of nodes, the system communication overhead and storage overhead are reduced. Security analysis proves that the scheme can resist tampering attacks, impersonation attacks, collusion attacks, double spend attacks, and replay attacks. Performance evaluation shows that the scheme improves compared to similar methods. Full article

Blockchain technology has rapidly evolved as a cornerstone of decentralized computing, transforming how trust, data integrity, and transparency are achieved in digital ecosystems. However, despite extensive adoption, significant gaps remain in understanding how key blockchain variables, such as block size, consensus mechanisms, and network latency, affect system vulnerabilities and susceptibility to cyberattacks. This survey addresses this gap by combining qualitative and quantitative analyses across multiple blockchain environments. Using simulation tools such as Ganache and Bitcoin Core, and reviewing peer-reviewed studies from 2016 to 2024, the research systematically maps blockchain parameters to cyberattack vectors including 51% attacks, Sybil attacks, and double-spending. Findings indicate that design choices like block size, block interval, and consensus type substantially influence resilience against attacks. The Blockchain Variable Quantitative Risk Framework (BVQRF) introduced here integrates NIST’s cybersecurity principles with quantitative scoring to assess risks. This framework represents a novel contribution by operationalizing theoretical security constructs into actionable evaluation metrics, enabling predictive modeling and adaptive risk mitigation strategies for blockchain systems. Full article

Background/Objectives: The structure and adequacy of health financing critically shape population health outcomes. This study examines financing typologies in relation to healthy life expectancy (HALE) and infant mortality across 38 OECD countries and Türkiye (2000–2021), quantifying financing model effectiveness and sex-disaggregated disparities. Methods: Time-weighted averages (exponential weighting, λ = 1.5) emphasized recent policy environments while preserving historical context. Principal component analysis addressed multicollinearity among six financial indicators. Multidimensional scaling (stress = 1.16 × 10⁻¹²) and K-means clustering identified four financing typologies. TOPSIS composite scores measured proximity to ideal outcomes (maximum HALE, minimum infant mortality), with success rates calculated as the percentage achieving top-quartile performance (TOPSIS ≥ 70). Sex-disaggregated analysis examined gender gaps across clusters. Results: High-Public-Spending systems achieved an 81.2% success rate (mean TOPSIS = 76.0), those with Balanced High-Expenditure achieved 77.8%, whereas Moderate/Emerging systems exhibited only 8.3% success. Türkiye ranked 36th of the 38 (TOPSIS = 24.8), 45% below cluster average, with extreme deficits in HALE (percentile = 15.8%) and infant mortality (7.9%). Low-resource systems showed significantly wider gender gaps (HALE: 3.43 vs. 1.66 years; infant mortality male excess: 1.04 vs. 0.53 per 1000; p < 0.01), with Türkiye demonstrating the third-highest male excess mortality globally (1.69 per 1000), indicating critical neonatal care deficiencies. Conclusions: Robust public financing (>USD 3500 per capita, >7% GDP) is necessary and nearly sufficient for superior outcomes, with success rates differing 10-fold between high- and low-resource systems (81% vs. 8%). Türkiye’s extreme underperformance reflects both inadequate public investment (USD 813 per capita, 22% of high-performing averages) and efficiency deficits requiring doubled expenditure alongside targeted maternal–child health interventions. Full article

Objectives: The aim of this study is to assess the impact of screen time on the incidence of sleep bruxism in children during the COVID-19 pandemic. Methods: The parents of 266 children, aged 3–14 years, participated in the present study. They were provided with a 36-item questionnaire in order to collect data about their child’s personal information, general health, sleep bruxism, and the effects of the COVID-19 pandemic on them. The collected data were analyzed statistically using a chi-square (χ²) test, ANOVA with post hoc analysis (Tukey’s HSD), and a t-test. Results: Screen time increased significantly during the pandemic, especially among children using screens ≥180 min/day. The proportion of children spending 180–360 min/day doubled to 24.4%. Lower secondary school children had the highest screen time, with an increase of ~60 min/day during the pandemic. Smartphones were the most used device (50.8%), and on average, children with bruxism spent 32 min longer in front of screens than children without bruxism (p < 0.05). Conclusions: Daily screen use is common in children, and this increased during the COVID-19 pandemic. Children with sleep bruxism exhibit longer screen time than those without bruxism, suggesting that the former is a potential risk factor for the latter. Full article

Background: The adoption of a healthy lifestyle and eating habits in children represents a major public health objective worldwide, with significant implications for the development of chronic non-communicable diseases in adulthood. In Italy, the “OKkio alla SALUTE” Surveillance System (National Institute of Health) has been in place since 2007 to periodically monitor the nutritional status and health-related behaviors of children aged 8 to 9 years old. Methods: Data were collected as part of the 2023 nutritional surveillance survey in the Tuscany Region through questionnaires completed by both children and their parents. A cluster sample design was adopted. The weight and height of children were directly measured. Logistic regression analysis was used to examine the association between measured variables (unhealthy eating habits and lifestyles) and overweight or obesity. Results: A total of 1427 children participated. In our sample, 17% of children were overweight, 5.7% were obese, and 1.3% were severely obese, totaling 24% of children classified as overweight. Tuscany’s rates are lower than the national average of 28.8%. Children whose parents had a low level of education were nearly twice as likely to consume sugary drinks daily (OR_adj = 1.97; 95% CI: 1.22–3.18) and to lead a sedentary lifestyle (OR_adj = 1.99; 95% CI: 1.33–2.97). Children from families reporting financial hardship were more likely to consume fruit and vegetables less than once a day (OR_adj = 2.35; 95% CI: 1.12–4.92) and to spend more time in sedentary activities (OR_adj = 3.30; 95% CI: 1.66–6.56). Regarding overweight, including obesity, children from economically challenged families had nearly double the risk of being overweight compared to those from financially stable households (OR_adj = 1.81; 95% CI: 1.09–2.98). Conclusions: The aim of our study was to evaluate which family factors are associated with unhealthy lifestyles in order to assess and, if appropriate, confirm the need for targeted and integrated interventions involving families, schools, and local communities to promote healthy lifestyles and effectively combat childhood obesity in Tuscany. Full article

Off-chain payments in the Internet of Things (IoT) enhance the efficiency and scalability of blockchain transactions. However, existing privacy mechanisms face challenges, such as the disclosure of payment channels and transaction traceability. Additionally, the rise of quantum computing threatens traditional public key cryptography, making the development of post-quantum secure methods for privacy protection essential. This paper proposes a post-quantum ring signature scheme based on hash functions that can be applied to off-chain payments, enhancing both anonymity and linkability. The scheme is designed to resist quantum attacks through the use of hash-based signatures and to prevent double spending via its linkable properties. Furthermore, the paper introduces an improved Hash Time-Locked Contract (HTLC) that incorporates a Signature of Knowledge (SOK) to conceal the payment path and strengthen privacy protection. Security analysis and experimental evaluations demonstrate that the system strikes a favorable balance between privacy, computational efficiency, and security. Notably, the efficiency benefits of basic signature verification are particularly evident, offering new insights into privacy protection for post-quantum secure blockchain. Full article