NMTMD (NMT-Melinda-Dataset)

Official repository for the Opensource Text dataset for NMT for local languages in West Africa (EWE Corpus) and implement the Yodi model afterward.

Note: This repository will evolve into the official repository for the Yodi model, once the necessary data is gathered.

Objective

• Develop a Machine Translation Text and Speech Dataset NMT for local languages in West Africa (EWE Corpus)

Key Results

-> Develop & Measure the accuracy or the performance of the Yodi model from this dataset for text-to-text translation.

-> Develop & Measure the accuracy of Yodi built from this dataset for Speech recognition.

Remark: Getting accurate data and labeled data from available sources online or in local written papers would be necessary for machine sentence translation.

Using the Updated Dictionaries

We have transformed and analyzed two Ewe-English dictionaries: KABDICT525 and EWEDICT995. These dictionaries are now available as Python modules for easy integration into your projects.

Accessing the Dictionaries

1. The transformed dictionaries are located in the Dictionaries folder:

   • Dictionaries/kabdict525.json
   • Dictionaries/ewedict995.json

2. To use these dictionaries in your Python scripts, you can import them as follows:

import json

# Load KABDICT525
with open('Dictionaries/kabdict525.json', 'r', encoding='utf-8') as f:
    kabdict = json.load(f)

# Load EWEDICT995
with open('Dictionaries/ewedict995.json', 'r', encoding='utf-8') as f:
    ewedict = json.load(f)

# Example usage
print(kabdict.get('word', 'Word not found'))
print(ewedict.get('word', 'Word not found'))

Ongoing

Feel free to share your analytics in the discussions! Instructions are available in project_contributions_instructions.txt

Please register at https://sites.google.com/umbaji.org/yodi/home to build the largest NMT text Dataset for West Africa.