English RobertaForTokenClassification Cased model (from obi)

Description

Pretrained RobertaForTokenClassification model, adapted from Hugging Face and curated to provide scalability and production-readiness using Spark NLP. deid_roberta_i2b2 is a English model originally trained by obi.

Predicted Entities

DATE, L-AGE, U-PATIENT, L-STAFF, U-OTHERPHI, U-ID, EMAIL, U-LOC, L-HOSP, L-PATIENT, PATIENT, PHONE, U-PHONE, L-OTHERPHI, HOSP, L-PATORG, AGE, U-EMAIL, L-ID, U-HOSP, U-AGE, OTHERPHI, LOC, ID, U-DATE, L-DATE, U-PATORG, L-PHONE, STAFF, L-EMAIL, PATORG, U-STAFF, L-LOC

Live Demo Open in Colab Download Copy S3 URI

How to use

documentAssembler = DocumentAssembler() \
        .setInputCol("text") \
        .setOutputCol("document")

sentenceDetector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx")\
       .setInputCols(["document"])\
       .setOutputCol("sentence")

tokenizer = Tokenizer() \
    .setInputCols("sentence") \
    .setOutputCol("token")

tokenClassifier = BertForTokenClassification.pretrained("roberta_ner_deid_roberta_i2b2","en") \
    .setInputCols(["sentence", "token"]) \
    .setOutputCol("ner")

pipeline = Pipeline(stages=[documentAssembler, sentenceDetector, tokenizer, tokenClassifier])

data = spark.createDataFrame([["PUT YOUR STRING HERE"]]).toDF("text")

result = pipeline.fit(data).transform(data)

val documentAssembler = new DocumentAssembler() 
          .setInputCol("text") 
          .setOutputCol("document")

val sentenceDetector = SentenceDetectorDLModel.pretrained("sentence_detector_dl", "xx")
       .setInputCols(Array("document"))
       .setOutputCol("sentence")

val tokenizer = new Tokenizer() 
    .setInputCols(Array("sentence"))
    .setOutputCol("token")

val tokenClassifier = BertForTokenClassification.pretrained("roberta_ner_deid_roberta_i2b2","en") 
    .setInputCols(Array("sentence", "token")) 
    .setOutputCol("ner")

val pipeline = new Pipeline().setStages(Array(documentAssembler,sentenceDetector, tokenizer, tokenClassifier))

val data = Seq("PUT YOUR STRING HERE").toDF("text")

val result = pipeline.fit(data).transform(data)

import nlu
nlu.load("en.ner.roberta.by_obi").predict("""PUT YOUR STRING HERE""")

Model Information

References

• https://huggingface.co/obi/deid_roberta_i2b2
• https://arxiv.org/pdf/1907.11692.pdf
• https://github.com/obi-ml-public/ehr_deidentification/tree/master/steps/train
• https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4978170/
• https://www.hhs.gov/hipaa/for-professionals/privacy/laws-regulations/index.html
• https://github.com/obi-ml-public/ehr_deidentification
• https://github.com/obi-ml-public/ehr_deidentification/tree/master/steps/forward_pass
• https://github.com/obi-ml-public/ehr_deidentification/blob/master/AnnotationGuidelines.md