Natural Language Processing Libraries

Question

I'm having a hard time figuring out what library and datasets go together.

Toolkits / Libraries I've found:

• CoreNLP - Java
• NLTK - Python
• OpenNLP - Java
• ClearNLP - Java

Out of all of these, some are missing features. For example OpenNLP didn't have a dependency parsing.

I need to find a library that's quick that will also do dependency parsing and part of speech tagging.

The next hurdle is where do we get data sets. I've found a lot of things out there, but nothing full and comprehensive.

Data I've found:

• NLTK Corpora
• English Web Treebank (looks to be the best but is paid)
• OpenNLP
• Penn Treebank

I'm confused as to what data sets I need for what features and what's actually available publicly. From my research is seems ClearNLP will work best for but has very little data.

Thank you

Answer 1

Stanford CoreNLP provides both POS tagging and dependency parsing out of the box (plus many other features!), it already has trained models so you don't need any data sets for it work!

Please let me know if you have any more questions about the toolkit!

http://nlp.stanford.edu/software/corenlp.shtml