YashvirGaming/mauritian-creole-english-corpus

🚀 4,610+ unique Mauritian Creole (Morisyen) to English parallel translation pairs. Structured JSON dataset featuring everyday phrases, slang, and cultural idioms with context. Built by Yashvir Gaming for NLP and LLM training. # Mauritian Creole to English Parallel Corpus (v1.0) This dataset contains **4610 unique, curated parallel translation pairs** mapping Mauritian Creole (*Morisyen*) to English. Unlike standard literal dictionaries, this corpus focuses heavily on everyday spoken phrases, cultural idioms (*proverb kreol*), internet meme syntax, and casual phonetic text-speak. Each entry contains a dedicated context field explaining the underlying cultural nuance, making it highly valuable for training Natural Language Processing (NLP) models in low-resource regional languages. ## Dataset Structure The dataset is provided as a single, unified JSON array of objects. Each object follows this exact schema: ```json [ { "kreol_phrase": "Anou kas poz kot la plaz.", "english_translation": "Let's hang out at the beach.", "context": "Casual suggestion among friends to relax." } ] ``` ## Methodology This dataset was engineered by Yashvir Gaming. Subscribe on youtube.com The raw outputs were then programmatically merged, regex-extracted, and strictly deduplicated to ensure 100% unique textual pairs. ## License This dataset is licensed under the **MIT License**. You are free to use, modify, and distribute this data for both personal and commercial AI applications, provided that original attribution is maintained.