Java NLP - Insight from Text Data

Question

We would like to do a POC that uses Java based NLP libraries like Stanford Core NLP and/or Deeplearning4J to train/use models that can extract insight / meaning / summary and provide answers to user queries against a custom document repository / DB (NoSQL DB / CSV). Is there any good documentation/simple code examples that illustrates this capability of these Java NLP libraries?

Answer 1

Yes you can use Java for NPL, especially if you want to be faster than Python. The best toolkit you can use is OpenNLP Apache, with this you can do a pipeline just like Spacy in Python.

Answer 2

(DL4J maintainer here):

Deeplearning4j would be useful if you want to directly use transformers and other models. If you are just looking for an out of the box toolkit with models I would suggest stanford nlp and other similar frameworks.

Answer 3

I think the best choice is Apache openNLP

Here is the complete project documentation and here is a working example You can find hundreds of sample projects like this on GitHub