Hi,
As some of you would already know, I have chosen to add a command line suggestion feature to GNU Octave and I have decided to use Artificial Neural Networks to do the same. If someone had been following my e-mails, they would know that I have been trying to contact the people behind
octave-online.net to collect training data, which would essentially be a list of common typographic errors.
I am writing this email to inform the community that
octave-online.net has shared a list of 'top 1000 misspellings' with me via email and have agreed to share more data soon. This should serve as a starting point for this project.
The performance of this feature could be improved if the members of the community agree to share their list of typographic errors. I have an idea of a m-script that should be able to catch these typographic errors but I need the help of the community to understand how these errors could be reported back. Help and suggestions would be appreciated.
P Sudeepam