Share your research data

MentalDistress is a human preference dataset developed for evaluating personalized alignment in Natural Language Processing(NLP). This dataset is developed to support mental health text classification research in the Bangla language. It consists of manually curated and annotated English text samples categorized into five psychological states. The dataset is designed to facilitate supervised learning approaches for detecting emotional distress and high-risk mental health indicators from textual data. The corpus contains a total of 10,098 annotated text samples, distributed across five classes representing different mental health conditions. # Class Distribution The dataset includes the following categories: - Suicidal: 2,170 samples - Depressed: 2,050 samples - Anxious: 2,038 samples - Frustrated: 2,062 samples - Others: 1,778 samples The class distribution is relatively balanced, making the dataset suitable for multi-class classification experiments without severe class imbalance issues. # Data Collection and Annotation The text samples were collected from publicly available sources and manually reviewed. Each instance was carefully annotated according to predefined psychological category guidelines to ensure labeling consistency. Quality control measures were applied to maintain annotation reliability. # Key Features - Five-class mental health categorization - Manually annotated dataset - Fairly balanced class distribution - Suitable for classical ML and deep learning models - Supports low-resource Bangla NLP research # Potential Use Cases - Mental health text classification - Early detection of psychological distress - Bangla NLP research - Transformer-based model benchmarking - AI-assisted mental health screening research # File Format The dataset is provided in CSV format with the following columns: - Text: English textual content - Label: Class name (Anxious, Depressed, Frustrated, Others, Suicidal)

To evaluate online Mindfulness-Based Cognitive Therapy (MBCT) as an adjuvant treatment for psoriasis, focusing on severity of lesion, anxiety, depression, quality of life, and itching. This randomized trial enrolled 109 psoriasis patients assigned to either treatment as usual (TAU) or TAU plus MBCT, which included eight weekly online sessions. Primary outcomes, included Psoriasis Area and Severity Index (PASI), Self-Rating Anxiety/Depression Scale (SAS/SDS), and Dermatology Life Quality Index (DLQI). Itching was a secondary outcome measured by Visual Analogue Scale (VAS). Assessments were conducted at baseline, 4, 8, and 12 weeks. Treatment effects were analyzed with mixed linear models. 91 patients completed the full follow-up: 46 in the MBCT+TAU group and 45 in the TAU group. Significant group × time interactions favoring MBCT+TAU were observed for PASI, SDS, DLQI and VAS. No significant between-group difference was found for anxiety.

This repository provides the full extraction dataset and quality appraisal outputs underpinning a PRISMA-guided systematic review of digital and Quality 4.0 technologies in construction quality management. The final dataset synthesises 51 included studies published between 2006 and 2026 and captures how technologies are being applied across the quality hierarchy (Inspection, Control, Assurance, and Management), with particular attention to adoption and governance conditions. For each included study, the dataset records bibliographic metadata (title, authors, year, and country/region as reported), research method and sample characteristics, and the primary “Application Level” within quality management. It documents the “Topic/technology investigated” and associated enabling infrastructures (e.g., AI/ML and computer vision, IoT/sensing, robotics and UAV-enabled inspection, blockchain-based traceability and e-inspection, BIM/digital twin and cloud/platform monitoring, and text mining/NLP of defect records). In addition, it captures reported benefits and drawbacks, reception/implementation context, referenced frameworks, and all KPI/metric reporting (algorithm-level metrics, process/outcome indicators, and whether tool metrics are linked to at least one process/outcome measure), including any reported accuracy/performance values where available. To support interpretation, the repository also includes Supplementary Table 2 reporting study quality appraisal using the Mixed Methods Appraisal Tool (MMAT), including Q1–Q5 criterion coding and overall ratings for all 51 studies. The dataset is intended to be reusable for secondary synthesis, benchmarking, and routine-centred evaluation of digital QA/QC workflows. It enables readers to trace where evidence is concentrated (inspection/control), where assurance/management implementations remain thinner, how KPI practices vary in clarity and comparability, and where governance-oriented technologies (e.g., traceability/auditability) are being operationalised.

This dataset contains the supplementary materials associated with the study “Comparative Hazards of Ocular Surface Disorders with IL-4/13 versus IL-13 Inhibitors in Atopic Dermatitis: A TriNetX Database Study.” The materials include detailed coding definitions, additional analyses, and supporting figures that complement the main manuscript and provide transparency for the analytic methods and results.

This dataset is linked to the study "Visitor characteristics and museum fatigue: a case study at the ETRU museum in Rome". It includes behavioural (position tracking) and physiological (heart rate) data from an experiment investigating the role of individual characteristics on museum fatigue. The dataset is divided into five files: ETRU_DATA_RAW: Full dataset with all collected data, including unprocessed variables and excluded participants ETRU_R_PROJECT: The R-Studio project with the script used for all the data analysis ETRU_DATASHEET_CATEGORIES: Demographic characteristics of participants (Age, Sex, Education, Public). The R-Studio project uses this file for cluster analysis. ETRU_DATASHEET_LONG: Preprocessed dataset in long format excluding participants not included in the analysis. The R-Studio project uses this file for spatial time-series analysis. ETRU_DATASHEET_WIDE: Preprocessed dataset in wide format excluding participants not included in the analysis. The R-Studio project uses this file for correlation analysis. These files enable replication of the analyses in the main paper and provide additional opportunities for secondary analyses.

Supporting Information for "Quantitative analysis of Calabrian planktic foraminifer assemblages and paleoecology of the Eastern Mediterranean from the onshore epibathyal sedimentary archives (Mersin, Turkey)"

This dataset provides model outputs, summary tables, and reproducible scripts supporting an evaluation of baseline RUSLE (LS) versus deformation-aware LS′ derived from PSInSAR, and the downstream implications for routed sediment redistribution using TLSD in two dryland watersheds (Yarkon-Ayalon and HaBsor). Outputs are provided for (i) gross detachment estimates from RUSLE and (ii) routed net erosion/deposition and sediment export derived from TLSD, enabling scenario comparison between DEM-based LS and PSInSAR-conditioned LS′. Included materials comprise (1) geospatial rasters (GeoTIFF) at the analysis resolution used in the study (e.g., LS and LS′ surfaces, gross erosion rate, TLSD deposition and net balance, sediment export, and SDR outputs), (2) PSInSAR spatial-support masks and associated summaries used to document where LS′ reverts to LS outside PS-supported areas, (3) block-aggregated land-use statistics based on non-overlapping 10×10 pixel blocks (modal land-use class; ≥70% valid coverage) and non-parametric comparisons (Kruskal-Wallis with ε²; Mann-Whitney with Holm correction; Cliff’s δ), and (4) Python scripts (rasterio + matplotlib) used to generate the main-paper and supplementary figures and tables. All p-values are reported descriptively given spatial dependence; effect sizes are provided to support interpretation. PSInSAR deformation is used to condition LS′ and does not directly measure erosion. Where third-party input datasets cannot be redistributed, the repository includes derived products and scripts/instructions to reproduce results from the original sources.

All model files (baseline and optimized), together with the corresponding Genetic Algorithm scripts and their associated input and output datasets, are provided to ensure transparency and reproducibility of the reported results.

This study experimentally evaluates the effectiveness of RFID technology for self-checkout applications in the Fast-Moving Consumer Goods (FMCG) sector, aiming to identify the key factors affecting reading accuracy. An experimental campaign was conducted using a fractional Design of Experiments (DoE) approach, considering shopping cart material, product quantity, presence of liquid products, presence of metal-containing products, and use of shielding bags as independent variables. Reading accuracy was analyzed through ANOVA to quantify the influence of physical and operational factors. The results show that reading accuracy ranges from 73% to 100%. The presence of liquid products emerges as the most significant factor, causing the largest performance degradation. The impact of metal-containing products, although difficult to isolate due to the pervasive presence of metal in carts and shielding bags, is statistically significant and mainly results in an increased performance variability. Product quantity exhibits a negligible effect. The highest accuracy is achieved under favorable configurations, particularly in the absence of liquids. Significant interaction effects are also observed, especially between cart material and the presence of metals. The study provides a quantitative assessment of RFID self-checkout performance in realistic FMCG environments, highlighting its current technical limitations, and offering practical insights for technology developers and retailers.

Long-term, repeated exposure to heat stress may contribute to chronic kidney disease (CKD), yet global evidence remains limited. In this ecological study, we aimed to examine the association of different categories of heat stress, as well as the overall heat stress score, with the incidence of CKD at the country/region level. Data from 174 countries/regions were analyzed. Annual numbers of days with moderate, strong, very strong, and extreme heat stress in each country/region between 1990 and 2021 were assessed using the Universal Thermal Climate Index. A weighted heat stress score was generated from these counts. Linear mixed-effects models were used to evaluate the association between heat stress score and CKD incidence. The results showed that all heat stress categories were positively associated with CKD incidence. In models using the weighted heat stress score, each standard deviation increase in the score was associated with an estimated 0.890 (95% CI: 0.717–1.063) additional incident cases per 100,000 population. In this ecological study, we observed country/region-level associations between heat stress and CKD incidence; nevertheless, further validation is warranted.