Is there a way to deidentify websites? right now it looks like the pretrained ner_deid_large NER model doesn't catch them Spark NLP for Healthcare

[egenc]

websites, we sometimes have texts in the end of our reports saying where to sent feedback and the website addresss has identifying information about the hospital

[egenc]

I have solved it checking this link:
https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tuto[…]reprocessing_with_SparkNLP_Annotators_Transformers.ipynb
https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tuto[…]gs/Healthcare/1.2.Contextual_Parser_Rule_Based_NER.ipynb