Loading...
Loading...
Student performance, MOOC logs, knowledge tracing, standardized tests, learning analytics
13,882 datasets
51,760 entries of instruction-following conversation data, originally created by Stanford University for training language models. This version has been cleaned, fixed, and converted into the ShareGPT format. The dataset was last updated on HuggingFace by SicariusSicariiStuff in November 2025.
Graph-structured benchmarks for node classification and link prediction were curated by Benedek Rozemberczki, with the latest update in December 2025. The repository provides data for evaluating algorithms in network science and machine learning research.
Delaware school district boundaries published by data.delaware.gov. The dataset includes columns such as NAME, ENROLL, SQMILES, SHAPE.AREA, and the_geom, suggesting it contains administrative and geographic information. It was last updated on 2025-11-01.
Jackvial released this repository in December 2025 as an unofficial editor for robotics datasets compatible with the LeRobot library. It facilitates the management of data for behavioral cloning and embodied AI, though specific record counts are not provided in the source. The tool is designed to handle the specialized data structures required for training robotic imitation learning models.
A research database containing results from a bibliometric analysis and systematic review of scientific output on digital pedagogy in higher education. The dataset was created by Santoveña Casal, Sonia and was last updated on November 23, 2025. The analysis focuses on studies concerning digitization, university education, and educational quality.
A validation procedure for a training system to teach ScratchJr to children, applied with 50 participants aged 6-7. The dataset, harvested from e-cienciaDatos, was last updated on 2025-11-04. It evaluates the system's usability for both neurotypical and neurodivergent children according to their preferences and needs.
1,000 medical case reports and associated question-answer pairs designed for medical residency training. The dataset supports the development of retrieval-augmented generation systems by providing complex clinical scenarios and expert-curated educational content.
Example Nsfw Dating Img is a dataset uploaded to Hugging Face by author PetruZetta. The dataset was last updated on December 12, 2025. Its specific content, scale, and structure require verification after download due to minimal provided metadata.
Chicago Public Schools provides official geospatial boundaries for high school attendance in the 2025-2026 academic year. The dataset includes polygon geometry and school identifiers, enabling precise mapping of student assignment zones. Its cross-platform presence on Socrata and Data.gov signals its importance for public planning and transparency.
Global flood detections derived from Sentinel-1 Synthetic Aperture Radar imagery using a deep learning change detection model. The dataset spans October 2014 to September 2024, offering a longitudinal view of flood-prone areas worldwide and accompanies a 2025 Nature Communications paper.
A Web Map Service (WMS) provides the original copy of a sub-local development plan for a school in the Velpke velvet community. The data originates from the Bundesamt für Kartographie und Geodäsie and was last updated on October 29, 2025. The service likely contains geospatial layers detailing zoning, infrastructure, or land use specific to the school area.
2,000 vision-text multimodal physics problems spanning from middle school to doctoral qualification exams. The SeePhys benchmark systematically evaluates LLMs/MLLMs on tasks integrating complex scientific diagrams with theoretical derivations. It was created by SeePhys and last updated on 2025-10-28.
30 interviews with patients from underserved minority groups explore experiences with serious illness conversations. Seiko Izumi conducted interviews averaging 38 minutes at four primary care clinics and one nursing home in the Pacific Northwest. The study aims to identify structures that make these clinical conversations accessible and acceptable to vulnerable populations.
106 policy and procedure documents from Auburn University, Ohio State University, and Virginia Tech were collected for dissertation research. The documents focus on disability and accessibility topics and are organized by institution and year. Martina Svyantek compiled this collection as part of a proof-of-concept project for QDR Harvested Dataverse.
Chea Stevenson's qualitative study explores correlates of sexual risk-taking behavior among emerging adults at a large public university in Coastal Kenya. Data were collected between October 31st 2019 and March 16th 2020 from 21 undergraduate students and 5 key informants, including the Dean of Students and health staff. Analysis employed a thematic framework approach and causal loop diagramming to map interconnected risk factors.
Scott W. H. Young collected survey and interview data on ethical library assessment practices in North America. The survey, conducted in November and December 2020, recorded 239 responses, including 73 complete submissions. The data includes survey documentation, codes, analysis, and interview protocols.
Eight performance-based tasks scored from video-recorded discussions of preservice elementary teachers with fifth-grade student avatars. Educational Testing Service and Mursion developed these tasks over a four-year project to improve teachers' ability to orchestrate discussions in science and mathematics. The data includes scoring materials and formative feedback reports generated from the project.
36 semi-structured interviews with Chinese consumers who purchased or needed customized furniture between February and April 2022. Researchers from Nanjing Forestry University developed a 4E interview-grounded theory model to explore demand factors across essential information, information extraction, user experience, and product expectation dimensions. The study included participants aged 20-55, with half male and half female respondents.
1,201 questionnaire responses and 175 simulation session records from a study examining how students judge online information credibility. The mixed-methods research by Tara Cataldo captured real-time judgments from K-12 and higher education students using simulated search engine results pages. Data includes think-aloud transcripts, task decisions, and click data from participants across six educational stages.
Conversations with a custom GPT assistant named 'GamifIcA Edu' were systematically validated across five distinct educational scenarios. The dataset likely contains the natural language interactions from four different tests per scenario, representing varied learning contexts across multiple academic disciplines. This work by López Galisteo, Antonio Julio was harvested on October 14, 2025.