FeniVerse: A Parallel Corpus of Feni Dialect, Standard Bengali, and English

Published: 11 September 2025| Version 1 | DOI: 10.17632/25r923frfj.1
Contributors:
,
,

Description

FeniVerse is an openly accessible trilingual parallel corpus containing 4,094 aligned sentences each in English, Standard Bangla, and the Feni dialect (12,282 total sentences). Each entry is sentence-aligned across the three languages, enabling machine translation, dialect classification, cross-linguistic analysis, and other NLP research. The dataset is provided as a ZIP file named “FeniVerse Parallel Corpus”, which contains two main files: FeniVerse_Dataset.csv – with three columns: English, Standard Bangla, Feni Dialect FeniVerse_Dataset.xlsx – with the same three columns This is the first publicly available dataset for the Feni dialect, offering an authentic, manually curated, and sentence-aligned resource for linguistic and computational experiments.

Files

Institutions

  • Daffodil International University

Categories

Natural Language Processing, Machine Translation, Machine Learning, Dialect, Bengali Language, Bangladesh, Deep Learning, Language

Licence