Advancing Open Science

Web accessibility assessment remains challenging because of the scale and heterogeneity of modern institutional websites and the difficulty of translating automatically detected violations into actionable remediation priorities. This paper introduces the Data-Driven Web Accessibility Remediation Framework (DD-WARF), a computational framework that integrates automated WCAG-based accessibility evaluation with statistical analysis, density-based clustering, similarity network analysis, and quantitative remediation modeling. DD-WARF combines DBSCAN-based anomaly detection to identify atypical accessibility profiles, cosine-similarity networks to characterize recurring patterns across institutions, and a Weighted Migration Effort Index (WMEI) to estimate relative remediation effort and support intervention prioritization. The framework was empirically evaluated using accessibility data from 320 Latin American university websites assessed with Google Lighthouse and WAVE under WCAG 2.2. The analysis revealed no statistically significant association between institutional ranking and the number of automatically detected accessibility errors ( , ). Three WCAG success criteria, 1.4.3 (Contrast (Minimum)), 1.1.1 (Non-text Content), and 2.4.4 (Link Purpose (In Context)), emerged as the dominant accessibility barriers, affecting nearly 90% of the evaluated institutions. DBSCAN further identified nine institutions with atypical accessibility profiles, while similarity network analysis revealed recurring accessibility patterns shared across institutions. By applying a simulated remediation scenario targeting these dominant criteria, the DD-WARF framework projects a theoretical accessibility improvement ranging from 37.4% to 70.6% across institutional profiles. These findings show the potential of DD-WARF as a reproducible, data-driven decision-support framework for large-scale accessibility diagnosis, benchmarking, remediation prioritization, and resource allocation.

Bicycle road infrastructure on weak or organic subgrades requires ground improvement or foundation solutions that vary significantly in material demand and construction complexity, yet their environmental implications remain poorly understood. This study applies a cradle-to-grave life cycle assessment (LCA), conducted in accordance with ISO 14040/44 and EN 15804+A2 using the Environmental Footprint (EF) 3.1 method, to compare three asphalt bicycle road designs differing solely in their substructure and foundation approach: a lightweight EPS geofoam embankment (Design 1), conventional non-cohesive soil replacement (Design 2), and a CFA micropiles system with a geogrid-reinforced working platform (Design 3). Ten midpoint impact categories were evaluated over a 15-year service life per 1 km of pavement. Results reveal that the choice of foundation solution is the decisive factor in environmental performance, with Design 1 achieving the lowest impacts across all ten indicators, followed by Design 3, and Design 2 performing worst despite being the most conventional approach. The environmental penalty of soil replacement stems from the disproportionate mass of embankment fill required (≈67,839 t versus ≈3955 t for the geofoam alternative), which amplifies burdens across material production, construction machinery use, and end-of-life processing. For the pile-based design, long-distance multimodal transport of imported aggregate dominates the impact profile, contributing over 54% of total GWP. This finding indicates that the environmental performance of the pile-supported alternative is highly sensitive to material sourcing and could be significantly improved through the use of locally available aggregates. The use and maintenance stage contributes identically and negligibly across all three designs, confirming that observed differences originate exclusively from substructure and foundation-related material flows. These findings demonstrate that substituting large soil volumes with lightweight engineered materials in pavement foundation design can deliver substantial environmental benefits without compromising structural function, offering a quantitative basis for integrating environmental criteria into foundation selection for cycling infrastructure on problematic subgrades.

Luminescence spectroscopy is an essential, non-destructive technique for mineralogical identification, yet conventional DIY spectrometers face noise, lossy compression, and low sensitivity when capturing weak Stokes-shifted emissions. To address these limitations, we developed a DIY spectrofluorometer that integrates a microscope objective and a transmission diffraction grating. The optical assembly is completed by a back-illuminated monochrome CMOS sensor (Sony IMX678). The camera is read by a local Python bridge through the manufacturer’s native SDK, which streams the uncompressed 12-bit sensor data to a custom web application (SpectraLib) via WebSockets, bypassing video-driver compression and 8-bit browser-canvas quantization. SpectraLib performs real-time ambient background subtraction and spectral smoothing. The detection capability of the instrument was illustrated by recording the weak liquid water Raman scatter band near 414 nm under 365 nm excitation (long-wave UV). The system characterized photoluminescence of six luminescent minerals from four hand specimens under either 365 nm or 255 nm excitation. It detected the Cr³⁺ R-line emission envelope in ruby at 693 nm, alongside the uranyl (UO²₂⁺) vibronic series in hyalite and chalcedony between 504 and 573 nm. It also resolved the broad radiation-induced yellow emission band in zircon centered near 600 nm. Additionally, the system resolved the characteristic orange-yellow emission of sodalite near 625 nm and the deep red emission of orthoclase near 690 nm, corresponding to sulfur-radical and trivalent iron activators, respectively. Emission spectra were visualized using real-time CIE 1931 chromaticity approximations, yielding uncorrected, instrument-dependent color estimates. Crystalline phases were verified via complementary Raman and FT-IR spectroscopies. With a total component cost of approximately $680, this open-architecture spectrofluorometer provides a web-based, high-dynamic-range alternative for educational and laboratory mineralogy, with prospective field applications.

Background/Objectives: Decision-making in advanced cancer care is intertwined with diagnostic profiling, prognostic communication, supportive care, and anticancer treatment selection. An author-developed D-S-T × B/C/E conceptual mapping scaffold—three decision/action domains (Diagnosis, Supportive care, Therapy) and three evaluative dimensions (Biological, Clinical, Existential)—was used to organise the literature, to identify coverage gaps and recurring framework categories, and, as a secondary conceptual synthesis objective, to derive a candidate integrative framework. Methods: The review was conducted according to the Joanna Briggs Institute (JBI) methodology for scoping reviews, pre-registered on OSF, and reported per PRISMA-ScR. PubMed was searched on 21 June 2026, and the final evidence set was derived solely from this search. Eligible records concerning adults with advanced cancer were dual-screened and charted using a structured form. Results: Eighty-two records (2010–2026) were included. Under the strict D-S-T × B/C/E matrix, Supportive care showed the highest joint coverage, particularly S × C (59/82, 72.0%), whereas D × B and D × E were the least represented cells (13/82 each, 15.9%). Eight recurring framework categories were inductively derived from the charted framework descriptors, led by early integrated palliative care (30/82) and prognosis-informed shared decision-making (27/82). Twenty-two records met an emergence-candidate screen defined at the synthesis stage, of which three received the maximum cross-axis coupling score of 6/6 from both reviewers. Conclusions: The advanced-cancer decision-framework literature clusters around a small number of architectures with uneven D-S-T × B/C/E coverage. Diagnosis-centred frameworks coupled with biological or existential dimensions were relatively under-represented within our mapping and represent priority targets for future primary research; the proposed Adaptive framework provides one structure in which this research could be conducted.

Japan’s hydrogen strategy faces three interrelated challenges in policy support, industrial deployment, and import security. This study builds a three-system framework covering policy, industry, and import security, and evaluates their coupling coordination in Japan from 2017 to 2025 using an entropy-weighted method and a coupling coordination degree model. A geopolitical impulse weighting factor based on Japan’s annual geopolitical risk index is further introduced to examine how external shocks reshape coordination. Results show moderate-to-high coupling but persistently low coordination. The policy and industry subsystems show partial co-movement in the earlier period but diverge more clearly after 2022, while the unweighted import-security subsystem also weakens after peaking around 2022. Geopolitical-risk adjustment further lowers coordination after 2022 without changing the overall temporal pattern. From a sustainability perspective, these findings suggest that Japan’s hydrogen transition requires better coordination among policy support, industrial development, and resilient supply structures. The proposed framework provides a transferable approach for assessing the coordination and sustainability of hydrogen transitions in energy-constrained economies.

The economic output of sea urchin aquaculture depends almost entirely on a single organ. Gonads, marketed as uni, determine value through their size, color, texture and flavor, and feed formulations have been optimized empirically for decades with limited reference to the molecular machinery governing gametogenesis, sex differentiation and gonadal maturation in echinoids. This review integrates four lines of mechanistic evidence around the adult gonad, namely developmental gene regulatory networks, sex-biased gonadal transcriptomics, germ-cell and nutritive-phagocyte biology, and endocrine, neuroendocrine and retinoid signaling, and reads them in terms of the control points connecting molecular state to production outcome. Rather than advancing a unified mechanistic model, which the current evidence does not support, each component is classified by the strength of the evidence attached to it: demonstrated in an echinoid, inferred from another taxon, plausible but untested, or hypothetical. The resulting picture is one in which descriptive resources have outpaced functional tests. Candidate sex-determination transcription factors have been identified but none have yet been directly perturbed in an echinoid, the retinoid pathway is architecturally present yet biochemically incomplete, and responses to exogenous steroids remain heterogeneous across species and protocols. Priority experimental directions are ranked by commercial return and by experimental falsifiability.

Reliable detection of small unmanned aerial vehicles (UAVs) is difficult in grayscale imagery degraded by low illumination, reduced contrast, noise, and blur. We propose stochastic degradation coverage training (SDCT), an offline strategy that broadens training data without changing the detector. A single-channel YOLOv8n model, denoted YOLOv8n-Gray, is trained using clean images and five degraded replicas per image. On the DUT Anti-UAV-derived benchmark, three-seed mean mAP50 on the seen degraded suite increases from 0.7369 with exposure-matched Clean Repeat to 0.8419. On the DUT-Adv target-domain test package, mAP50 is 0.8089 for Clean Repeat, 0.8519 for DART, and 0.8481 for SDCT. On FRED’s 47-sequence canonical test split, pooled mAP50 is 0.0062, 0.0052, and 0.0059, respectively, with Clean Repeat reporting the highest value. On the 13-sequence challenging intersection, SDCT exceeds Clean Repeat (0.0080 vs. 0.0073), while DART is highest (0.0088). Thus, FRED indicates a conditional gain over clean repetition rather than consistent SDCT superiority. Under a synthetic held-out stress test, SDCT achieves mean mAP50 of 0.6939. TensorRT FP16 inference on Jetson AGX Orin achieves 4.23 ms latency and 236.36 FPS. External test sets were used for final evaluation.

Ports are cyber-physical systems, and compromising them has physical consequences: paralysed cranes, blocked gates and cargo that does not move. Research on protecting them remains divided between cybersecurity engineering, port management and supply chain studies, with no shared unit of analysis. This review adopts the maritime port supply chain as that unit and asks which vulnerabilities and threat vectors are specific to its IT and operational technology (OT) environments, which regulatory instruments apply and how they interrelate, and which resilience strategies the literature supports, and on what kind of evidence. A combined bibliometric and qualitative review design was adopted to provide both a macro-level overview of the field and an in-depth examination of the technical and regulatory evidence. In total, 21,268 records were retrieved from Scopus and the Web of Science, of which 14,659 remained after deduplication and the application of a 2010 lower bound. The corpus was mapped with bibliometrix and VOSviewer and examined qualitatively alongside relevant regulatory instruments. Five thematic clusters emerge, showing a temporal shift from SCADA and infrastructure protection towards intrusion detection, IoT and machine learning. Exposure is cumulative: IT enables access, OT determines consequence, and software supply chains propagate both. The regulatory regime remains ship-biassed, with the main gap lying in implementation rather than in the absence of instruments. The controls most consistently recommended are architectural and organisational rather than novel, but only one of the thirteen appraised is supported by three or more independent sources generated in port or maritime settings. A framework for proactive port cyber resilience is proposed.