bolingual

A reproducible benchmark and retrieval system for finding English words that sound similar to Hindi words written in Devanagari.

Find English words that sound similar to Hindi words written in Devanagari. The package includes:

  • A curated benchmark of 2,959 Hindi-English pairs from Xlit-Crowd

  • Three retrieval methods: orthographic, phonetic, and hybrid

  • CLI tools and Python API for evaluation and querying

Contents

Indices