Hindi → Indian Languages Education Translation Dataset

Overview

Parallel corpus for Hindi → multiple Indian languages focused on the education domain.
Content spans textbooks-style explanations and texts across all subjects like History, Geography, Chemisty, Biology etc.
Each subfolder maps to a language pair (e.g., HIN-ASM, HIN-BAN, HIN-DOI/ODI, HIN-GOM, HIN-GUJ, HIN-KAS, HIN-MAI, HIN-MAR, HIN-MNI, HIN-NPI, HIN-PAN, HIN-SAT, HIN-SND, HIN-TEL; and Tamil-centric pairs TAM-KAN, TAM-MAL, TAM-TEL if included).

Languages

Source: Hindi (hi)
Targets: Assamese (as), Bangla (bn), Goan Konkani (gom), Gujarati (gu), Kashmiri (ks), Maithili (mai), Marathi (mr), Manipuri/Meitei (mni), Nepali (npi), Odia (or), Punjabi (pa), Santali (sat), Sindhi (sd), Telugu (te), Tamil (ta), Kannada (kn), Malayalam (ml), Urdu (ur).

Sources

Replace with exact attributions:

• School education department circulars and notices
• NCERT/SCERT materials and summaries (publicly available)
• University/board announcements and FAQs

Format

• Folder arrangement: Each language pair is subdivided into sub-domains and each sub-domain is further divided into source-translated, source_reviewed and translated-reviewed language pairs.
• Type folders: source-reviewed translations are the translations of the highest quality which are reviewed after translation; whereas source-translated language pairs are larger in number but are unreviewed after manual translation. The difference between the qualities can be estimated using the translated-reviewed language pairs using TER.
• File type: TSV per language pair
• Fields: id, src_hi, tgt_<lang>, domain="education"
• UTF-8 encoding, normalized punctuation

Intended Use

• Training & evaluation of MT systems for educational content in India.