roman-urdu-spell-correction

implementation of the paper "A Spelling Correction Program Based on a Noisy Channel Model"

This code was originally written by github.com/jbhoosreddy
I only modified it for Roman Urdu corpus

How to run

Download the corpus.data file from https://drive.google.com/file/d/12JUTse_jGhziTqIrGUCJdTjzkZ6lyKO0/view?usp=drive_link and paste it in the project root/main directory
Execute 'spellcorrect.py' file

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
.gitattributes		.gitattributes
README.md		README.md
addconfusion.data		addconfusion.data
delconfusion.data		delconfusion.data
dictionary.data		dictionary.data
format_data.ipynb		format_data.ipynb
ngram.py		ngram.py
revconfusion.data		revconfusion.data
spellcorrect.py		spellcorrect.py
spellerrors.data		spellerrors.data
subconfusion.data		subconfusion.data