Help needed - extracting number of incidences from text data

I am working on social media data wherein I need to come up with the total number of posts/tweets that mention that someone took a particular medicine/treatment by referring to various posts and tweets.
I am trying with NER using spacy to look for a particular pattern but it is not giving accurate results.
What approach can I take to count number of incidences where user took a particular medicine or treatment?

@vidulak4 ,
A reason why spacy might not be giving good results is because it isn’t trained specifically for medicine related terms.
So you will you have to train your own NER model which focuses on medicines/medical terms.
A simple google search for “NER model medicine” yields very good resources.
Some good resources - 1 , 2

Thank you for the input. Will check it out.

© Copyright 2013-2020 Analytics Vidhya