- Add option "remove_stopwords" to remove stop words from text during correction
- Add function to check correction candidates determined with the word embedding model against the lemma list
- Add new pretrained matching dict based on the improed embedding model routine
- Replace "tolist" parameter in normalizer function with "output" parameter (string, list, json)
- Output to json may include original word, correction, lemma, n-rule correction, correction index and text position index (depending on parameters)
- Performance improvements for lemmatizer function
- Add folder "workflows" to repository for example usage and application scripts
- Bug fixes and code cleanup
- Replace fuzzywuzzy with jellyfish for fuzzy string matching (using Jaro-Winkler distance)
- Add license info for norvig corrector
- Bug fixes and code cleanup
- First public development build