Datasets:
Bilingual Multimodal CS Competition Dataset
This dataset contains multiple-choice questions from MateInfoUB a high-level computer science competition, provided in both English and Romanian, with optional associated images and detailed image descriptions. The goal is to support research in educational technology, multilingual NLP, and automated evaluation systems for competitive programming.
Dataset Features
- 🗣 Languages: English (
en.json), Romanian (ro.json) - 🧠 Modality: Text-based questions, optional image with text description
- 📚 Format: Multiple-choice (single correct answer)
- 🧩 Fields:
title: Short question namemarkdown_statement: Full question body in Markdown formatanswer_variants: List of multiple-choice optionscorrect_answer: String index of the correct answerdifficulty: One of"easy","medium","hard"image_path(optional): Relative path to imageimage_content(optional): Textual description of the image for accessibility/vision tasks
- Downloads last month
- 3