license: gpl-3.0
task_categories:
- text-classification
- text-generation
- sentence-similarity
language:
- ru
pretty_name: Openedu
size_categories:
- 1K<n<10K
1.2k OpenEdu.ru courses (2025)
Overview
The dataset contains structured information about courses from the OpenEdu.ru platform: course cards with basic metadata, extended information extracted from each individual course page, and a mapping between course directions (groups). It is useful for analysing educational programmes, searching via metadata, and training text-classification models, including RuBERT and other transformers, because it includes direction codes that can serve as target labels. It can be used to train LLMs to predict students' skills based on course content.
Provenance
The dataset was collected from publicly available pages on the OpenEdu.ru platform in November 2025. Basic course information was extracted from the main catalogue using Selenium with Firefox and geckodriver to load dynamically generated content, and detailed course data were retrieved from individual course pages using Requests and BeautifulSoup with the lxml parser. User-Agent values were randomised during scraping, and parallel processing was performed with a thread pool. All text fields and dates were preserved in the exact format presented on the website.