Malayalam, the native language of Kerala, India is agglutinative. This makes it particularly challenging to write a spell checker for it. Currently, there is no open-source spell checker available.
This is a proof of concept for an algorithm for Malayalam spell checking. Given a Joiner that implements word compounding rules and a word, the Splitter splits the word into morphemes. These morphemes can then be checked against a dictionary.
- A Joiner that implements Malayalam compounding rules exhaustively need to be created
- Splitter should be tested against a large corpus to check if it is viable for all scenarios