Publicly Available Resources for Bangla-English-Bangla Machine Translation

Note: We listed the following tools and resources for the sake of their dissemination and accessibility. We neither claim their ownership nor taking any responsibility to their uses. Please use and cite the appropriate authors if you use them for your research work. If you use them with any of your software application please contact the authors OR use them at your own risk.

Parallel Corpus

1. Indic Languages Multilingual Corpus. Click here to Download
2. Six Indian Parallel Corpora
3. Penn Treebank Bangla-English Parallel Corpus
4. AmaderCAT Parallel Corpus

Tokenizer

1. Moses Tokenizer
2. Indic NLP Tokenizer
3. Bangla Tokenizer (coming soon...)

Machine Translation Training and Evaluation Tools

1. Moses Toolkit (Statistical Machine Translation)
2. OpenNMT Toolkit (Neural Machine Translation)
3. Google Seq2Seq Model (Neural Machine Translation)
4. NVIDIA Seq2Seq Model (Neural Machine Translation)
5. Harvard Seq2Seq Attention Model (Neural Machine Translation)

Machine Learning Framework

1. PyTorch
2. TensorFlow

Parallel Corpus Development Tools

1. AmaderCAT (Simplified and Collaborative)
2. OmegaT(Free Offline Platform)
3. Zanata (Open Source)
4. SDL Trados (Commercial)
5. Sketch Engine ((Commercial))