English Text to Triple (S->P->O) Generation with Fast AI

Hi All,

I have the following text in documents:

Document 1: Dr Steve West is the president of XYZ Holding that is located in Bristol, UK. He founded XYZ Holding in January 2012.
Document 2: Octavia Spencer, who is Dr Steve’s wife, was the director of XYZ Holding from January 2014 to December 2016.

And I want to create a classifier to generate triples from the above text using Fast AI deep learning algorithms:

Dr Steve West President of XYZ Holding
XYZ Holding Located in Bristol, UK.
XYZ Holding founded in January 2012
Octavia Spencer wife of Dr Steve West
And so on.

This is the first time I encountered this type of NLP task and don’t have much knowledge of how to best approach this task. Has anyone ever worked on such kind of problem using Fast AI or have seen someones work on a similar task? Can you pls share the links to such works?

If not, I will really appreciate any guidance or help on how to best tackle this problem using Fast AI.

Thanks in advance

Im new with FastAI but i would take a look at NLP tools like Spacy and verb phrases. Maybe FastAI uses this too?