A Parallel Dataset for Standard Bangla to Seven Regional Dialects

Description

This dataset is developed for Standard Bangla to regional Bangla dialect translation. It consists of 2,500 Standard Bangla source sentences collected from the Vashantor dataset, along with corresponding English and Banglish representations and translations into seven regional varieties: Pabna, Noakhali, Jashore, Rangpur, Mymensingh, Barishal, and Chittagong. Pabna, Noakhali, Jashore, and Rangpur contain 2,500 translated sentences each, while Mymensingh and Barishal contain 1,500 sentences each and Chittagong contains 1,130 sentences. Regional translations were produced and reviewed by native speakers to ensure semantic equivalence, dialect authenticity, spelling consistency, and grammatical quality. The dataset supports research in dialect translation, machine translation, multilingual NLP, and linguistic analysis of regional Bangla varieties.

Institutions

• Daffodil International University

  Dhaka Division, Dhaka

Categories

Licence