ArzEn-MultiGenre: A Comprehensive Parallel Dataset

Overview

ArzEn-MultiGenre is a distinctive parallel dataset that encompasses a diverse collection of Egyptian Arabic content. This collection includes song lyrics, novels, and TV show subtitles, all of which have been meticulously translated and aligned with their English counterparts. The dataset serves as an invaluable tool for various linguistic and computational applications. Published: 28 December 2023
Version: 3
DOI: 10.17632/6k97jty9xg.3
Contributor: Rania Al-Sabbagh

Dataset Details

• Total Segment Pairs: 25,557
• Languages: Egyptian Arabic and English
• Content Types: Song Lyrics, Novels, TV Show Subtitles

Applications

• Machine Translation Benchmarking: Ideal for testing and improving new machine translation models.
• Language Model Fine-Tuning: Suitable for enhancing large language models in few-shot settings.
• Commercial Application Adaptation: Can be used to refine tools like Google Translate for better performance with Egyptian Arabic.

Research Relevance

This dataset is a significant resource for research in fields such as translation studies, cross-linguistic analysis, and lexical semantics.

Unique Contributions

1. Diverse Textual Genres: The dataset includes genres not typically found in parallel datasets for Egyptian Arabic and English.
2. Gold-Standard Quality: Translated and aligned by human experts, ensuring high accuracy and reliability.

Citation

Please cite this dataset as follows:
Al-Sabbagh, Rania (2023). “ArzEn-MultiGenre: An aligned parallel dataset of Egyptian Arabic song lyrics, novels, and subtitles, with English translations.” Mendeley Data, V3, DOI: 10.17632/6k97jty9xg.3

Related Links

• Article

Institutions

• University of Sharjah