Search Results (720)

With the digital economy advancing at a fast pace, digital transformation plays a pivotal role in reinforcing firms’ innovation capability and promoting high-quality development. This study analyzes Chinese non-financial publicly listed firms on the A-share market over the period 2009–2023. Based on text mining of annual reports, this study constructs an index capturing digital transformation and empirically evaluate its impact on innovation output with firm and year fixed effects. The estimates suggest that digital transformation meaningfully increases firms’ innovation output; the inference is unchanged when applying instrumental-variable approaches and conducting extensive robustness checks. Mechanism analysis reveals two parallel channels: (1) the R&D investment mechanism, characterized by improvements in R&D intensity, capitalization rate, per capita efficiency, and investment growth; (2) the governance environment mechanism, reflected in enhanced internal control, improved information disclosure quality, and strengthened audit supervision. Once firms are stratified by characteristics, the estimated positive effect of digital transformation is most pronounced for firms with low financial constraints, large size, eastern locations, and state ownership. This study identifies both direct and indirect mechanisms linking digital transformation to innovation and highlights how firm- and region-specific features condition the magnitude of this effect, thereby offering empirical implications for corporate digitalization strategies and policy design. Full article

Prompt specifications for multi-agent large language model (LLM) systems carry data contracts and integration logic across interdependent files but are rarely subjected to structured-inspection rigor. We report a single-system case study of iterative, agent-driven auditing applied to AEGIS (Autonomous Engineering Governance and Intelligence System), a seven-lane production pipeline whose 7152-line specification surface was audited across nine rounds, surfacing 51 consistency defects (per-round counts of 15, 8, 12, 2, 8, 1, 4, 1, 0). We present a seven-category post hoc taxonomy with explicit coding rules, non-monotonic convergence consistent with cascading edits and audit-scope expansion, and a locked audit protocol. We further report two partial replications on a public synthetic mini-specification: a cross-LLM panel of four frontier vendors (OpenAI, Anthropic, Google, xAI; 12 traces; multi-vendor union detects all five seeded defects) and an inter-rater reliability check on a stratified subsample (Cohen’s $\kappa$ = 0.80 on category, 0.46 on severity). The full reproducibility bundle accompanies the submission. Full article

Natural gas metering is shifting from volume-based measurement to energy-based assessment as gas sources diversify, pipeline networks become more interconnected, and gas quality varies more strongly across time and space. This review examines the key technologies required for natural gas energy metering and evaluates how they support full-chain traceability from production to end use. The reviewed topics include flow measurement, gas composition analysis, calorific value determination, temperature-pressure compensation, state correction, uncertainty evaluation, intelligent data acquisition, and metrological traceability. The literature shows that individual technologies have advanced substantially. Ultrasonic flowmeters, rapid gas-quality sensing methods, dynamic calorific value allocation models, high-accuracy equations of state, and digital metering platforms have improved the technical basis of energy metering. However, these advances remain more mature at the level of individual links than at the level of the complete metering chain. Under multi-source supply, gas-quality fluctuation, hydrogen blending, and digitalized operation, the main challenge is to maintain consistency, uncertainty control, online verification, data credibility, and auditability across different metering stages. Future development should therefore focus on dynamic calorific value allocation, robust state correction under variable gas quality, full-chain uncertainty propagation, online verification, and secure data management for traceable natural gas energy metering. Full article

Failure Mode and Effects Analysis (FMEA) remains a fundamental risk management methodology in the automotive industry. This review provides a structured comparative analysis of the classical AIAG FMEA (4th edition, 2008) and the harmonized AIAG & VDA FMEA (1st edition, 2019) across Design (DFMEA), Process (PFMEA), and System (SFMEA) levels. Unlike conventional descriptive reviews, this study presents an integrative analytical synthesis that systematically evaluates methodological differences, decision-making logic, and structural transformations between the two frameworks. The analysis focuses on key developments, including the transition from Risk Priority Number (RPN) to Action Priority (AP), the introduction of a mandatory seven-step methodology, the formalization of structure–function–failure relationships, and enhanced traceability to downstream quality documentation such as Control Plans. The findings demonstrate that the harmonized framework represents a conceptual shift from a primarily scoring-based approach to a structured systems engineering methodology, improving consistency, completeness, and auditability of risk analysis. Particular emphasis is placed on the implications of AP-based prioritization, which alters traditional decision logic by preventing the suppression of safety-critical risks. The paper contributes to the literature by providing a comprehensive cross-level comparison (DFMEA–PFMEA–SFMEA) within a single analytical framework, identifying both strengths and limitations of the harmonized approach, and outlining its practical implications for industrial implementation. Future research directions include quantitative validation, application-based case studies, and integration with digital and AI-driven FMEA systems. Full article

Shannon entropy is the most common measure that one could use to check if a data source has random behaviour or not. A value close to the maximum is usually considered as evidence that the source is “random enough”. The present paper shows that this criterion alone is not enough. A deterministic logistic map driven at $r=3.9999$ reaches 94.97% of the Shannon maximum, yet it is fully predictable once we look at the built-in patterns: its permutation entropy drops to 77.01% of the maximum and its sample entropy falls to 0.67, against 2.33 for a high-quality pseudo-random generator (PRNG). Building on this observation, we combine four entropy measures—Shannon, Rényi, permutation, and sample—into a single diagnostic profile of the analyzed source. In order to validate our approach with practical, real life data, we test it on 2538 official draws of the Romanian Loto 6/49 lottery, recorded between August 1993 and April 2026. The lottery historical data set is very close to a high-quality PRNG (pseudo-random number generator) from the point of view of all four measures. We also observe that the entropy deficit of both the lottery and the PRNG decays as a power law with exponent $\alpha \approx -0.96$; in contrast, the logistic map sits at $\alpha \approx -0.07$. A Random Forest classifier trained only on the entropy profile reaches 78% accuracy on the analyzed four-way classification task (lottery, PRNG, logistic map, biased distribution), but scores 55.7% on the binary lottery-versus-PRNG task, consistent with chance. The method introduced in this study is domain-independent and applies directly to RNG certification, cryptographic key auditing, and any setting where structured pseudo-randomness has to be ruled out. Full article

Maintaining consistent quality performance of recycled aggregates is essential for their reliable use in construction applications. This study evaluated whether the regulatory revision permitting concurrent appointment of quality and environmental managers affected post-certification quality test performance within Korea’s recycled aggregate certification system. Extending a previous 2025 audit-based study, this research analyzed 311 certification-application-level follow-up quality test results obtained during the 2023 national post-certification management process. Statistical analyses, including chi-square tests, Fisher’s exact tests, odds ratio comparisons, and subgroup analyses, were conducted according to management structure, personnel change status, and recycled aggregate application type. The results showed that concurrent appointment and personnel changes were not associated with statistically significant deterioration in post-certification quality test performance. In contrast, the recycled aggregate application type showed substantially greater influence on pass/fail outcomes, with relatively higher failure risks observed in concrete and fine aggregate applications requiring stricter quality control conditions. Road construction and asphalt concrete applications generally maintained relatively stable pass rates regardless of management structure or personnel continuity conditions. The subgroup analyses additionally showed that concurrent appointment did not significantly increase failure risk within any recycled aggregate application category. These findings indicate that concurrent appointment did not significantly deteriorate actual post-certification quality performance within the analyzed national certification dataset. Full article

Reliance on standalone accuracy limits credible assessment of fundus-focused large vision–language models (LVLMs), as high scores often stem from linguistic shortcuts rather than real visual reasoning. This work develops the Cognitive Audit Framework (CAF), a four-module automated auditing pipeline that dissects model reasoning flaws: Visual–Linguistic Decoupling (textual dependency via modality ablation), Hierarchical Logical Consistency (lesion–diagnosis contradiction detection), Reasoning Fidelity Gap (chain-of-thought unfaithfulness scoring), and Contextual Robustness (positional bias under option permutation). Experiments on six 7B–31B LVLMs over FunBench reveal a notable gap between benchmark accuracy and reasoning quality: high accuracy coexists with measurable textual dependency, logical inconsistencies across diagnostic levels, limited chain-of-thought faithfulness, and non-trivial positional sensitivity. CAF serves as a reproducible complement to pure accuracy metrics for validating clinical competence of ophthalmic multimodal models. Full article

Precise dimensional conformity of precast concrete members is critical for structural performance and on-site assembly accuracy, yet conventional manual inspection remains labor-intensive and unable to scale with modern production-line throughput. Existing scan-vs-BIM approaches address geometric verification in principle but are constrained by manual registration dependencies, the absence of machine-readable IFC-linked tolerance criteria, and limited validation under real factory yard conditions. This study presents a production-scale automated shape quality control (SQC) framework that closes all three gaps simultaneously. A purpose-designed two-point target device enables fully automated, repeatable registration seed-point extraction. A formal IFC property-set-linked rule engine architecture—comprising entity extraction, deviation computation, rule interpretation, and pass/fail decision stages—replaces ad hoc script-based tolerance checking with an interoperable, auditable compliance pipeline. Factory-scale validation on precast arch segments (n = 10) and wall panels (n = 12) achieved registration RMSE of 1.25–1.95 mm, pass rates exceeding 91%, and a 37.1% reduction in inspection time versus manual methods (95% CI: 34.5–39.6%; p < 0.001; Cohen’s d = 3.89). Repeatability testing yielded ICC = 0.971 and Bland–Altman limits of agreement of [−0.45, +1.07] mm. The framework represents a substantive step toward fully digital, production-integrated quality management for industrialized precast construction. Full article

Background: Artificial Intelligence (AI) software in mammography is trained on a set of processed images and may be less effective when applied to images acquired on different systems or systems with different processing and/or acquisition settings. The aim of this work was to undertake a retrospective audit of a large number of mammography systems in the United Kingdom and identify the number of differences in image acquisition and processing factors. Methods: Images of the TORMAM phantom are acquired as part of the routine quality control programme. Data from the DICOM header of these images were extracted to provide a snapshot in time of the system configurations. A longitudinal audit of DICOM header data for all of the Hologic systems was tested by one medical physics department (MPD1) over 14 years. Results: We received results from 28 UK medical physics services for 498 systems. There were 7 different models of mammography systems, each with up to 7 different versions of acquisition workstation software. Each mammographic model had multiple image processing versions, including bespoke settings. The GE had two dose settings, while Siemens systems had a range of doses from 80% to 150% of the standard dose. In the longitudinal audit, there were between 2 and 6 software versions in concurrent use on the Hologic systems tested by MPD1. Conclusions: This study showed the heterogeneity of system setup across the UK in a single year, as well as changes to system setup over time. These differences may affect the outcomes of both AI and human readers. There are responsibilities on AI suppliers, mammography equipment manufacturers, breast-screening units, and medical physics services to ensure outcomes are not adversely affected by differences or changes in mammography equipment configurations. Full article

Artificial intelligence is reshaping warehouse work through algorithmic task allocation, scanner-based monitoring, KPI feedback, dynamic scheduling, and real-time performance control. Although these systems can improve coordination and operational visibility, they also create governance risks related to fairness, transparency, autonomy, privacy, workload pressure, trust, and employee resistance. This study develops a human-centered decision framework for prioritizing algorithmic management governance packages in third-party logistics (3PL) warehousing. The main contribution is to translate employee-level governance concerns into a scenario-sensitive decision model that helps managers select appropriate governance packages under different operational pressures. The study uses survey data from 380 warehouse employees to examine key psychological and behavioral mechanisms, including procedural fairness, transparency, system/information quality, autonomy, privacy concern, workload, trust, acceptance, and resistance/disengagement. These survey-supported constructs are then converted into six governance criteria: procedural fairness, transparency and contestability clarity, system and information quality, autonomy support, privacy boundary governance, and workload protection. A seven-expert panel evaluates five governance packages under three scenarios: peak season surge, labor shortage/high turnover, and audit pressure/compliance scrutiny. Methodologically, the framework combines Dynamic Multi-Facet Fuzzy Sets to capture membership, non-membership, hesitancy, engagement, and resistance; Bayesian Network weighting to reflect dependencies among governance criteria; and PCA-based ranking optimization to generate scenario-specific and robust rankings. Comparative validation with SAW and TOPSIS is also used to assess ranking consistency. The findings show that effective algorithmic management governance is not a fixed compliance solution. Transparency, workload protection, autonomy support, privacy boundary governance, and procedural fairness become more or less important depending on the operational scenario. A2, which combines transparency, workload protection, and autonomy support, emerges as the strongest robust package. A1 performs best under labor shortage/high turnover, while A3 performs best under audit pressure/compliance scrutiny. These results suggest that 3PL warehouses should adopt adaptive governance routines that combine explainability, contestability, workload safeguards, privacy boundaries, and employee voice mechanisms. The study contributes to the literature on AI in socio-technical systems by showing how human, organizational, and ethical concerns can be embedded into an interpretable decision framework for responsible algorithmic management in logistics work environments. Full article

This research aims to assess the degree of risk disclosure practices in the Jordanian corporate sector and to examine the influence of critical audit quality dimensions—specifically audit opinion, audit fees, and auditor type—on the amount of corporate risk disclosure (CRD). The research examines 90 annual reports from Jordanian non-financial publicly traded companies from 2014 to 2023, resulting in 900 firm-year observations. A manual content analysis method was used to quantitatively assess the degree of risk disclosure, supplemented by logistic regression to analyze the influence of audit quality indicators. The empirical data indicate that the quantity of risk disclosure statements differs across businesses, spanning from 2 to 10 words, with a mean of 24 sentences. Furthermore, the findings indicate that characteristics influencing audit quality—specifically unqualified audit opinions, elevated audit fees, and Big Four auditors—exhibit a positive and substantial correlation with increased levels of risk disclosure. This indicates that enhanced audit quality elevates the legitimacy and openness of company reporting, thus reducing information asymmetry between management and stakeholders. Previous research on risk disclosure in Jordan has mostly neglected the influence of audit quality as a factor in transparency. This study is among the few that thoroughly investigate the impact of audit opinion, audit fees, and auditor type on company risk disclosure within the non-financial sector. The results underscore the essential importance of audit quality in improving monitoring and disclosure processes, thereby enriching the existing literature on corporate governance and risk reporting in developing economies. Full article

Multivariate time-series anomaly detection on physical sensor networks faces three challenges that generic deep learning models inadequately addressed: heterogeneous sensor reliability, context-dependent anomaly scoring, and inactionable binary outputs lacking per sensor attribution. We propose an uncertainty-aware Temporal Convolutional Network (TCN) framework built on two tightly integrated uncertainty-driven components: (i) an Adaptive Uncertainty-Aware Attention (AUAA) mechanism that gates temporal attention weights by per sensor predictive uncertainty obtained from Monte Carlo dropout; and (ii) a Dynamic Weight Adapter that learns context-sensitive blending of reconstruction error and uncertainty via a GRU over weight history. The architecture also includes an exploratory per sensor attribution head, which we audit rather than claim: a controlled-perturbation test shows it is not yet causally faithful. We complement the empirical architecture with two distribution-free theoretical results: a Chebyshev-type false-positive bound on the hybrid anomaly score, and a Monte Carlo posterior moment convergence result at rate $O\left(M^{-1/2}\right)$. Evaluated on four-month indoor air quality sensor data, the Full Enhanced model achieves $R^2=0.9988$ and MSE $1.65\times {10}^{-4}$, a $25.2\%$ MSE reduction over the Base TCN ($R^2=0.9984$, MSE $2.20\times {10}^{-4}$). Because the IAQ stream is unlabeled, the primary quantitative detection evaluation uses the labeled Skoltech Anomaly Benchmark (SKAB), a publicly available industrial water-circulation corpus disjoint from the IAQ training distribution; it yields an 8.8 × F1 advantage ($0.477$ vs. $0.054$) and a 14.4 × recall advantage ($0.418$ vs. $0.029$) for the proposed model configuration over the Base TCN at a validation-calibrated threshold applied without retuning. Against twelve established detectors under a unified protocol, the proposed model attains the best F1 and recall, while the strongest reconstruction baselines retain higher precision and a marginally higher ROC-AUC, a recall-driven trade-off. Ablation isolates each component’s contribution, the detector degrades gracefully under channel masking and noise, and the distribution-free false-positive bound is empirically respected. The framework retains a low inference cost ($0.16$ ms per window at $M=20$ Monte Carlo samples, including the uncertainty pass). Full article

In China, policy-driven rural renewal projects have transformed many village public spaces, but some renewed sites are still weakly integrated into villagers’ everyday routines. This study asks why some renewed public spaces sustain routine use and low-intensity social interaction, while others remain materially complete but socially weak. The study was conducted in a rural township in Puyang County, Henan Province. Twelve renewed public spaces across several villages were examined through structured spatial audits and 579 resident questionnaires. Five spatial conditions were assessed: visibility, stay support, activity accommodation, interaction-supportive arrangement, and experienced locational convenience. Two behavioral outcomes were used to describe post-renewal vitality: use frequency and social participation. The analysis combines necessary condition analysis (NCA) and fuzzy-set qualitative comparative analysis (fsQCA). NCA is used as a diagnostic tool for identifying upper-limit constraints, while fsQCA is used to identify sufficient combinations of conditions. The results suggest that experienced locational convenience is the clearest bottleneck condition for both outcomes. When a site is difficult to incorporate into residents’ daily walking routines, internal design quality has limited capacity to translate into sustained behavioral use. Among better-located spaces, high vitality is associated with several design configurations. The most stable recurrent pattern combines visibility, stay support, and locational convenience as core conditions, together with either interaction-supportive arrangement or activity accommodation. Low-vitality spaces follow a different logic, being characterized by the simultaneous absence of several supporting conditions rather than by the absence of one isolated feature. The paper therefore proposes a two-step diagnostic logic for rural public-space renewal: first checking whether a site is embedded in everyday mobility and then matching internal spatial conditions with local patterns of use. Full article

This paper provides a theory-informed synthesis of earnings management research through a review of the 100 most cited studies in the accounting literature. Rather than functioning as a purely bibliometric review, the study integrates theoretical, empirical, methodological, and survey-based contributions to examine how influential research has conceptualized, measured, and interpreted earnings management. Citation data were collected from Web of Science and Google Scholar as of 5 January 2025 using predefined search criteria, filtering procedures, and classification protocols. While citation counts are used to identify influential studies, they are not treated as direct indicators of research quality due to concerns regarding citation bias, publication visibility, and proxy limitations. The review organizes the literature around major themes, including corporate governance, audit quality, managerial incentives, institutional environments, market reactions, and regulatory change. The analysis highlights enduring debates concerning proxy validity, endogeneity and identification challenges, the distinction between statistical detection and economic significance, and the trade-off between accrual-based and real earnings management. The synthesis also incorporates emerging research streams involving family firms, gender diversity, ESG reporting, textual analysis, and AI-assisted analytics within broader agency and institutional theory perspectives. A central contribution of the paper is the development of an integrative analytical framework linking proxy validity, strategic substitution between reporting mechanisms, and institutional constraints within a unified interpretation of earnings management behavior. The review shows that advances in empirical design, textual analysis, machine learning, and predictive analytics extend rather than replace foundational insights, while persistent limitations in causal inference and measurement remain unresolved. Overall, the findings suggest that earnings management is best understood as a strategic response to incentives, monitoring, and institutional constraints rather than as a uniform indicator of opportunistic behavior. The paper concludes by outlining future research directions focused on theory-driven empirical design, methodological triangulation, AI-assisted detection approaches, and improved measurement frameworks across diverse reporting environments. Full article

Accurate viral protein function annotation is essential for genomic surveillance, yet conventional retrieval-augmented generation (RAG) pipelines often fragment biological evidence into fixed-length text chunks, disrupting relationships among ORFs, annotations, structural domains, sequence motifs, residue mappings, and model-derived attention evidence. We propose ViroBioTree, a tree-structured biological evidence retrieval framework for downstream viral protein evidence review rather than a new primary annotation classifier. Built as an evidence organization layer on ViralMultiNet-derived ORF-level predictions and annotations, ViroBioTree converts sequence, annotation, structure, and attention evidence into typed biological nodes and traceable edges, then performs deterministic multi-channel recall, evidence-aware reranking, balanced TopK selection, rule-based verification, and node-cited report generation. In a demo benchmark, ViroBioTree achieved its strongest deterministic proxy performance on structure-explanation tasks, with Precision@K = 1.0, Recall@K = 1.0, and diversity = 0.52; these values reflect expected node-type and tag agreement rather than independent biological correctness. A bounded full-scale SARS-CoV-2 index contained 39,800 ORF rows, 80,000 attention records, 199,418 nodes, and 495,886 edges. In a stratified full20k diagnostic evaluation, ViroBioTree showed task-dependent advantages over LlamaIndex vector retrieval for conflict detection, evidence retrieval, and structure explanation, while LlamaIndex remained competitive or stronger for annotation-rich function annotation. A cross-family Influenza A Virus (IAV) diagnostic audit showed that the schema can represent IAV evidence namespaces while explicitly exposing missing formal ORF inputs, missing attention evidence, and unavailable residue/PDB assertions. Supplementary robustness, external sanity-check, diversity-risk, expert-evaluation, domain-tool positioning, and cross-family audit analyses supported traceability, report quality, and conservative evidence handling, but also showed that stable Precision@K under query perturbation does not necessarily imply stable retrieved evidence sets. ViroBioTree operates offline and deterministically, but does not address raw-read assembly, base calling, primary ORF prediction, or wet-lab validation. Its results should be interpreted as proxy and expert-reviewed evidence for traceable viral protein evidence retrieval and report generation rather than as direct validation of biological function annotation. Full article