We have told you how to use nltk wordnet lemmatizer in python: Dive Into NLTK, Part IV: Stemming and Lemmatization , and implemented it in our Text Analysis API: NLTK Wordnet Lemmatizer. We have preprocessed the english text with pos tagger and then lemmatize them one by one.

Recently, one of our api user send us email with a special need:

I am currently subscribed to the TextAnalysis API through Mashape.com. It’s been working well. I am interested in using the lemmatizer API, but, I don’t see a way to specify that I want to lemmatize verbs at the endpoint (either https://textanalysis.p.mashape.com/mbsp-word-lemmatize or https://textanalysis.p.mashape.com/nltk-wordnet-lemmatizer). Is there an undocumented parameter that I should use? If I want to lemmatize verbs, how would I go about that?

This is a good suggestion, so this weekend I add a special english word lemmatize demo and api for english word only: NLTK Wordnet Word Lemmatizer. User can specified the pos tag which nltk wordnet lemmatizer support, only four tags can be used: v(verb), a(adjective), r(adverb), n(noun). Just enjoy it:

Demo: http://textanalysisonline.com/nltk-wordnet-word-lemmatizer
Api: https://www.mashape.com/textanalysis/textanalysis#nltk-wordnet-word-lemmatizer

Posted by TextMiner