User 2032 | 8/12/2015, 10:20:42 AM
There are some things that should be quite straight forward to do and could give you an age over other text analytics libraries.
First please include stopwords for different languages by default:
You could use these https://code.google.com/p/stop-words/ (free of charge)
Second it would be very cool if you would have pre-trained text feature extractors based on Word2Vec (same as with your recent awesome release of transfer learning for images) for different languages. This idea was inspired by the cool results found here: https://code.google.com/p/stop-words/