Information Extraction with Neural Networks and Free Noisy Supervision (arxiv.org)

How does the pipeline compare to state-of-the-art information extraction suites (e.g. Apache OpenNLP) on standard datasets?

Well. That was a short paper.

Where's the code?