-
-
Notifications
You must be signed in to change notification settings - Fork 19
Home
(English translation made by google translate.)
I aimed to develop utilities for Turkish natural language processing (ddi) with trnlp. I hope I can help you with your project. For trnlp, it would be better to say ddi tools rather than a library. Briefly, if we list what can be done with trnlp;
- Finding the word base and stem (Base, stem)
- Word and rule based morphological analysis (Lemmatization)
- Suggestion list for words written using wrong or Ascii characters (Spelling Corrector)
- Text shredding and counting operations (Tokenization)
trnlp needs a dictionary for each transaction. When you download trnlp, it will come with these dictionaries. Currently, there is no feature to change or add to dictionaries. All dictionaries are saved as .pickle files and are accessible. I think of adding the data adding feature to the dictionary in the next version. However, manual changes can be made, subject to the required dictionary format.
Detailed descriptions for each section are available on the relevant wiki page.