Edit on GitHub
  • John Snow Labs
  • Sep 10, 2025

Detect Adverse Drug Events (BertForTokenClassification - ONNX)

  • medical
  • clinical
  • ner
  • en
  • licensed
  • onnx

Description

Detect adverse reactions of drugs in reviews, tweets, and medical text using the pretrained NER model. This model is trained with the BertForTokenClassification method from the transformers library and imported into Spark NLP

Predicted Entities

I-ADE, B-DRUG, B-ADE, O, I-DRUG, PAD

Copy S3 URI

How to use

from sparknlp.base import DocumentAssembler
from sparknlp_jsl.annotator import MedicalBertForTokenClassifier
from sparknlp.annotator import Tokenizer, NerConverter
from pyspark.ml import Pipeline

document_assembler = (
    DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")
)

tokenizer = (
    Tokenizer()
    .setInputCols(["document"])
    .setOutputCol("token")
)

token_classifier = (
    MedicalBertForTokenClassifier.pretrained(
        "bert_token_classifier_ner_ade_onnx",
        "en",
        "clinical/models"
    )
    .setInputCols(["token", "document"])
    .setOutputCol("ner")
    .setCaseSensitive(True)
)

ner_converter = (
     NerConverterInternal()
    .setInputCols(["document", "token", "ner"])
    .setOutputCol("ner_chunk")
)

pipeline = Pipeline(stages=[
    document_assembler,
    tokenizer,
    token_classifier,
    ner_converter
])

test_sentence = "I have an allergic reaction to vancomycin so I have itchy skin, sore throat/burning/itching, numbness of tongue and gums. I would not recommend this drug to anyone, especially since I have never had such an adverse reaction to any other medication."
data = spark.createDataFrame([[test_sentence]]).toDF("text")

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

from johnsnowlabs import nlp, medical

document_assembler = nlp.DocumentAssembler()\
    .setInputCol("text")\
    .setOutputCol("document")


tokenizer = nlp.Tokenizer()\
    .setInputCols(["document"])\
    .setOutputCol("token")


token_classifier = medical.BertForTokenClassifier.pretrained(
        "bert_token_classifier_ner_ade_onnx",
        "en",
        "clinical/models"
    )\
    .setInputCols(["token", "document"])\
    .setOutputCol("ner")\
    .setCaseSensitive(True)


ner_converter = medical.NerConverterInternal()\
    .setInputCols(["document", "token", "ner"])\
    .setOutputCol("ner_chunk")


pipeline = Pipeline(stages=[
    document_assembler,
    tokenizer,
    token_classifier,
    ner_converter
])

test_sentence = "I have an allergic reaction to vancomycin so I have itchy skin, sore throat/burning/itching, numbness of tongue and gums. I would not recommend this drug to anyone, especially since I have never had such an adverse reaction to any other medication."
data = spark.createDataFrame([[test_sentence]]).toDF("text")

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

import com.johnsnowlabs.nlp.base.DocumentAssembler
import com.johnsnowlabs.nlp.annotators.Tokenizer
import com.johnsnowlabs.nlp.annotators.ner.NerConverter
import com.johnsnowlabs.nlp.annotators.classifier.dl.MedicalBertForTokenClassifier
import org.apache.spark.ml.Pipeline

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

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

val tokenClassifier = MedicalBertForTokenClassifier
  .pretrained("bert_token_classifier_ner_ade_onnx", "en", "clinical/models")
  .setInputCols(Array("token", "document"))
  .setOutputCol("ner")
  .setCaseSensitive(true)

val nerConverter = new  NerConverterInternal()
  .setInputCols(Array("document", "token", "ner"))
  .setOutputCol("ner_chunk")

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

val testSentence = "I have an allergic reaction to vancomycin so I have itchy skin, sore throat/burning/itching, numbness of tongue and gums. I would not recommend this drug to anyone, especially since I have never had such an adverse reaction to any other medication."
val data = Seq(testSentence).toDF("text")

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

Results


+---------------------------+------+
|text                       |entity|
+---------------------------+------+
|allergic reaction          |ADE   |
|vancomycin                 |DRUG  |
|itchy skin                 |ADE   |
|sore throat/burning/itching|ADE   |
|numbness of tongue and gums|ADE   |
+---------------------------+------+

Model Information

Model Name: bert_token_classifier_ner_ade_onnx
Compatibility: Healthcare NLP 6.1.1+
License: Licensed
Edition: Official
Input Labels: [document, token]
Output Labels: [ner]
Language: en
Size: 403.7 MB
Case sensitive: true
Max sentence length: 128