SPARKNLP application in R

[ghost]

Hi,
I'm reading the paper about implementing NER in SPARK NLP for Clinical Research, 'Improving Clinical Document Understanding on COVID-19 Research with SPARK NLP'. There are examples of how to implement the NLP pipeline at the back of the paper. I'm using R so following David Kincade's SPARK NLP
R implementations at https://github.com/r-spark/sparknlp/commit/156aea018c37f4fb082399f2d5ee5810c199ade6

What I can't figure out is how to use one of the pretrained models e.g. ner_diseases within this framework.

I presume it's the following block of code that implements this but not sure where to load model.
Do we need to preload language models like spacy? Any guidance? Has anyone else implemented these pipelines in R?

clinical_ner = NERDLModel.pretrained("
      ner_clinical", "en", "clinical/models")
      \
      setInputCols(["sentence", "token", "
      embeddings"])\
      .setOutputCol("ner")

[maziyarpanahi]

Hi,

Anything that says clinical/models is a licensed model/pipeline and they also have their own annotators which only work with licensed keys. You can ask for trial keys or an academic license.

In any case, this repository is only for public/open-source products/models. If you have more questions you can ask in Slack under #healthcare channel, they answer pretty quickly.

[ghost]

Thank you