While using the deidentification models on Spark NLP for Healthcare, is there a way to tell it to obfuscate the ids? #310
Answered
by
JustHeroo
JustHeroo
asked this question in
sparknlp-healthcare
-
|
In the deidentification models, is there a way to tell it to obfuscate the ids? currently on obfuscation mode, I'm getting whenever it finds one, I would like it to be some random ID (or even not so random) instead. |
Beta Was this translation helpful? Give feedback.
Answered by
JustHeroo
Aug 25, 2021
Replies: 1 comment
-
|
Sure, it is called Obfuscation mode. You can use faker to get a random ID or set your own list of fake ids to be used. Please check this notebook >> https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tuto[…]n_Trainings/Healthcare/4.Clinical_DeIdentification.ipynb |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
JustHeroo
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Sure, it is called Obfuscation mode. You can use faker to get a random ID or set your own list of fake ids to be used.
Please check this notebook >> https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tuto[…]n_Trainings/Healthcare/4.Clinical_DeIdentification.ipynb