Loading...
Loading...
Student performance, MOOC logs, knowledge tracing, standardized tests, learning analytics
13,874 datasets
FinePDFs-Edu dataset consists of over 350 billion tokens of educational text extracted from PDFs. The dataset was created by Web3Survivor using an educational quality classifier trained on annotations from a large language model to filter content from the broader FinePDFs collection. It covers 69 languages and was last updated on December 11, -0001.
A dataset of 2,200 prompts designed to elicit deep reasoning across math, coding, precise instruction following, and general chat. It was created by the mlx-community to train the Olmo-3-7B-Think model and was last updated on November 21, 2025. The dataset blends high-quality curated sources with filtering for deliberate reasoning and is converted for compatibility with MLX-LM-LoRA.
Washington state data estimates childcare demand and supply gaps for children from infancy through age 12. The Department of Children, Youth, and Families (DCYF) estimates families of about 293,000 pre-school children and over 491,000 school-age children need care, with licensed or subsidized services meeting only 27% and 12% of that need respectively. The dataset was last updated in September 2025.
Information compiled for the CIVE3207 (ARCN4100) Historic Site Recording and Assessment course in 2025. Undergraduate students of the Architectural Conservation and Sustainability Program at Carleton University will use these data to produce a Heritage Recording of this important Site. The dataset was authored by Adam Weigert and last updated on December 27, 2025.
1,309 passenger records from the 1912 shipwreck categorized by survival status and demographic details. The data includes 12 distinct variables such as passenger class, age, and family relations to support binary classification modeling.
University Medical Center Groningen data likely contains medical or clinical research information. The dataset is hosted on the Papers with Code platform, which suggests a connection to computational research. Specific details about its size, format, and creation date are unknown.
Cleanlab is an open-source library for data-centric AI and machine learning quality, developed by the Cleanlab organization and updated through January 2026. It provides a standardized package for identifying and fixing issues in messy, real-world datasets and labels.
Featuring the survey instruments used in a 2022 research project at Rotterdam University of Applied Sciences to measure perceived data literacy among business programme students and lecturers. The questionnaires operationalize data literacy across six core competencies: awareness, access, engagement, management, communication, and use. The instruments include full item lists, Likert-scale response structures, and demographic questions.
A global collection of approximately 7,000 hours of real-world robot data from FrodoBots EarthRover units. The dataset, created by frodobots and last updated in November 2025, is a filtered and cleaned subset of a larger 10k-hour collection, reannotated using the Model-Based Reannotation (MBRA) method to improve action label fidelity.
Japanese literary texts sourced from the Aozora Bunko digital library, enhanced with jReadability-based difficulty analysis. The dataset was created by ronantakizawa and last updated on November 19, 2025. It is intended for Japanese language learning and curriculum development.
AllenAI provides a dataset of synthetic question-and-answer pairs generated using the Gemini 2.5 Flash model. The dataset contains 417,000 entries intended for research and educational purposes. It was last updated on December 16, 2025.
MSHA (Mine Safety and Health Administration) data on assessed violations linked to individual case numbers. The dataset is maintained by the U.S. Department of Labor and was last updated in January 2026. Specific details on row count, columns, and temporal coverage are not provided in the input.
Cardiff University translation memories provide a collection of English-Welsh sentence pairs. The dataset is intended for tasks including language modeling, parsing, and sentiment analysis. It was created by techiaith and last updated on November 28, 2025.
UniPercept is a benchmark dataset for unified perceptual-level image understanding across aesthetics, quality, structure, and texture. Developed by thunderbolt215 and updated in February 2026, it provides a multi-dimensional evaluation framework for multimodal large language models (MLLMs).
Synthetic question-and-answer pairs generated by the Gemini 2.5 Flash model. The dataset is provided by AllenAI under the ODC-BY 1.0 license and was last updated in December 2025. It is intended for research and educational use.
RAGCare-QA is a benchmark dataset of 420 theoretical medical knowledge questions for evaluating Retrieval-Augmented Generation pipelines. The dataset was created by ChatMED-Project and was last updated on November 20, 2025. Each question is annotated with its optimal RAG pipeline type.
ivrit.ai's Crowd Recital dataset is designed for training Whisper models on Hebrew audio. The dataset was last updated on 2025-11-19. It contains audio data, likely paired with text, for speech-related tasks.
Synthetic question-and-answer pairs generated by the Gemini 2.5 Flash model. The dataset is licensed under ODC-BY 1.0 and intended for research and educational use by AllenAI. It was last updated on December 16, 2025.
A study of 176 future education professionals in Spain analyzed the influence of three digital pedagogical models—collaborative, social, and independent—on learning experience. The dataset contains results from a Likert-scale questionnaire with 20 closed and 9 open questions, administered during the 2019-20 academic year. Sonia Santoveña Casal authored this research, which was harvested by e-cienciaDatos.
A curated collection of Web3 question-answer pairs for training and evaluating domain-specific AI models. The dataset includes metadata fields for category and difficulty. It was created by author 'iioos' and last updated on December 15, 2025.