Description
Zero-shot Named Entity Recognition (NER) enables the identification of entities in text with minimal effort. By leveraging pre-trained language models and contextual understanding, zero-shot NER extends entity recognition capabilities to new domains and languages. While the model card includes default labels as examples, it is important to highlight that users are not limited to these labels. The model is designed to support any set of entity labels, allowing users to adapt it to their specific use cases. For best results, it is recommended to use labels that are conceptually similar to the provided defaults.
Predicted Entities
AdmissionDischarge, Age, Allergen, BodyPart, ClinicalDept, DateTime, Disease,
Dosage, Drug, Duration, Employment, Form, Frequency, Gender, Laterality, MedicalDevice,
Modifier, Procedure, PsychologicalCondition, RaceEthnicity, Substance, Symptom, Test,
Treatment, Vaccine
How to use
document_assembler = DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentence_detector = SentenceDetector()\
.setInputCols(["document"])\
.setOutputCol("sentence")
tokenizer = Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
labels = [
'AdmissionDischarge', 'Age', 'Allergen','BodyPart','ClinicalDept','DateTime','Disease',
'Dosage','Drug','Duration','Employment','Form','Frequency','Gender','Laterality','MedicalDevice',
'Modifier','Procedure','PsychologicalCondition','RaceEthnicity','Substance','Symptom','Test',
'Treatment','Vaccine']
pretrained_zero_shot_ner = PretrainedZeroShotNER().pretrained("zeroshot_ner_vop_large", "en", "clinical/models")\
.setInputCols("sentence", "token")\
.setOutputCol("ner")\
.setPredictionThreshold(0.5)\
.setLabels(labels)
ner_converter = NerConverterInternal()\
.setInputCols("sentence", "token", "ner")\
.setOutputCol("ner_chunk")
pipeline = Pipeline().setStages([
document_assembler,
sentence_detector,
tokenizer,
pretrained_zero_shot_ner,
ner_converter
])
data = spark.createDataFrame([["""Hello,I'm 20 year old girl. I'm diagnosed with hyperthyroid 1 month ago. I was feeling weak, light headed,poor digestion, panic attacks, depression, left chest pain, increased heart rate, rapidly weight loss, from 4 months. Because of this, I stayed in the hospital and just discharged from hospital. I had many other blood tests, brain mri, ultrasound scan, endoscopy because of some dumb doctors bcs they were not able to diagnose actual problem. Finally I got an appointment with a homeopathy doctor finally he find that i was suffering from hyperthyroid and my TSH was 0.15 T3 and T4 is normal . Also i have b12 deficiency and vitamin D deficiency so I'm taking weekly supplement of vitamin D and 1000 mcg b12 daily. I'm taking homeopathy medicine for 40 days and took 2nd test after 30 days. My TSH is 0.5 now. I feel a little bit relief from weakness and depression but I'm facing with 2 new problem from last week that is breathtaking problem and very rapid heartrate. I just want to know if i should start allopathy medicine or homeopathy is okay? Bcs i heard that thyroid take time to start recover. So please let me know if both of medicines take same time. Because some of my friends advising me to start allopathy and never take a chance as i can develop some serious problems.Sorry for my poor english😐Thank you."""]]).toDF("text")
result = pipeline.fit(data).transform(data)
document_assembler = nlp.DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentence_detector = nlp.SentenceDetector()\
.setInputCols(["document"])\
.setOutputCol("sentence")
tokenizer = nlp.Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
labels = [
'AdmissionDischarge', 'Age', 'Allergen','BodyPart','ClinicalDept','DateTime','Disease',
'Dosage','Drug','Duration','Employment','Form','Frequency','Gender','Laterality','MedicalDevice',
'Modifier','Procedure','PsychologicalCondition','RaceEthnicity','Substance','Symptom','Test',
'Treatment','Vaccine']
pretrained_zero_shot_ner = medical.PretrainedZeroShotNER().pretrained("zeroshot_ner_vop_large", "en", "clinical/models")\
.setInputCols("sentence", "token")\
.setOutputCol("ner")\
.setPredictionThreshold(0.5)\
.setLabels(labels)
ner_converter = medical.NerConverterInternal()\
.setInputCols("sentence", "token", "ner")\
.setOutputCol("ner_chunk")
pipeline = nlp.Pipeline().setStages([
document_assembler,
sentence_detector,
tokenizer,
pretrained_zero_shot_ner,
ner_converter
])
data = spark.createDataFrame([["""Hello,I'm 20 year old girl. I'm diagnosed with hyperthyroid 1 month ago. I was feeling weak, light headed,poor digestion, panic attacks, depression, left chest pain, increased heart rate, rapidly weight loss, from 4 months. Because of this, I stayed in the hospital and just discharged from hospital. I had many other blood tests, brain mri, ultrasound scan, endoscopy because of some dumb doctors bcs they were not able to diagnose actual problem. Finally I got an appointment with a homeopathy doctor finally he find that i was suffering from hyperthyroid and my TSH was 0.15 T3 and T4 is normal . Also i have b12 deficiency and vitamin D deficiency so I'm taking weekly supplement of vitamin D and 1000 mcg b12 daily. I'm taking homeopathy medicine for 40 days and took 2nd test after 30 days. My TSH is 0.5 now. I feel a little bit relief from weakness and depression but I'm facing with 2 new problem from last week that is breathtaking problem and very rapid heartrate. I just want to know if i should start allopathy medicine or homeopathy is okay? Bcs i heard that thyroid take time to start recover. So please let me know if both of medicines take same time. Because some of my friends advising me to start allopathy and never take a chance as i can develop some serious problems.Sorry for my poor english😐Thank you."""]]).toDF("text")
result = pipeline.fit(data).transform(data)
val document_assembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val sentence_detector = new SentenceDetector()
.setInputCols("document")
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols("sentence")
.setOutputCol("token")
labels = Array(
"AdmissionDischarge", "Age", "Allergen", "BodyPart", "ClinicalDept", "DateTime", "Disease",
"Dosage", "Drug", "Duration", "Employment", "Form", "Frequency", "Gender", "Laterality", "MedicalDevice",
"Modifier", "Procedure", "PsychologicalCondition", "RaceEthnicity", "Substance", "Symptom", "Test",
"Treatment", "Vaccine")
val pretrained_zero_shot_ner = PretrainedZeroShotNER().pretrained("zeroshot_ner_vop_large", "en", "clinical/models")
.setInputCols(Array("sentence", "token"))
.setOutputCol("ner")
.setPredictionThreshold(0.5)
.setLabels(labels)
val ner_converter = new NerConverterInternal()
.setInputCols(Array("sentence", "token", "ner"))
.setOutputCol("ner_chunk")
val pipeline = new Pipeline().setStages(Array(
document_assembler,
sentence_detector,
tokenizer,
pretrained_zero_shot_ner,
ner_converter
))
val data = Seq([["""Hello,I'm 20 year old girl. I'm diagnosed with hyperthyroid 1 month ago. I was feeling weak, light headed,poor digestion, panic attacks, depression, left chest pain, increased heart rate, rapidly weight loss, from 4 months. Because of this, I stayed in the hospital and just discharged from hospital. I had many other blood tests, brain mri, ultrasound scan, endoscopy because of some dumb doctors bcs they were not able to diagnose actual problem. Finally I got an appointment with a homeopathy doctor finally he find that i was suffering from hyperthyroid and my TSH was 0.15 T3 and T4 is normal . Also i have b12 deficiency and vitamin D deficiency so I'm taking weekly supplement of vitamin D and 1000 mcg b12 daily. I'm taking homeopathy medicine for 40 days and took 2nd test after 30 days. My TSH is 0.5 now. I feel a little bit relief from weakness and depression but I'm facing with 2 new problem from last week that is breathtaking problem and very rapid heartrate. I just want to know if i should start allopathy medicine or homeopathy is okay? Bcs i heard that thyroid take time to start recover. So please let me know if both of medicines take same time. Because some of my friends advising me to start allopathy and never take a chance as i can develop some serious problems.Sorry for my poor english😐Thank you."""]]).toDF("text")
val result = pipeline.fit(data).transform(data)
Results
+--------------------+-----+----+----------------------+----------+
|chunk |begin|end |ner_label |confidence|
+--------------------+-----+----+----------------------+----------+
|20 year old |11 |21 |Age |0.99549705|
|girl |23 |26 |Gender |0.99473166|
|hyperthyroid |48 |59 |Disease |0.72869056|
|1 month ago |61 |71 |DateTime |0.93013126|
|weak |88 |91 |Symptom |0.98246646|
|light |94 |98 |Symptom |0.914502 |
|digestion |112 |120 |Symptom |0.7636672 |
|panic attacks |123 |135 |PsychologicalCondition|0.7021233 |
|depression |138 |147 |PsychologicalCondition|0.88090825|
|left |150 |153 |Laterality |0.9534648 |
|chest |155 |159 |BodyPart |0.90008795|
|pain |161 |164 |Symptom |0.96790904|
|heart rate |177 |186 |Test |0.86297375|
|weight loss |197 |207 |Symptom |0.96102 |
|4 months |216 |223 |Duration |0.9639058 |
|hospital |259 |266 |ClinicalDept |0.93731934|
|discharged |277 |286 |AdmissionDischarge |0.9054539 |
|hospital |293 |300 |ClinicalDept |0.93180996|
|blood tests |320 |330 |Test |0.94237006|
|brain |333 |337 |BodyPart |0.7430142 |
|mri |339 |341 |Test |0.93738127|
|ultrasound scan |344 |358 |Test |0.94679016|
|endoscopy |361 |369 |Test |0.5370005 |
|doctors |392 |398 |Employment |0.7688288 |
|homeopathy doctor |487 |503 |Employment |0.93872774|
|he |513 |514 |Gender |0.974668 |
|hyperthyroid |547 |558 |Disease |0.71458954|
|TSH |567 |569 |Test |0.9895893 |
|T3 |580 |581 |Test |0.9048731 |
|T4 |587 |588 |Test |0.9588674 |
|b12 deficiency |614 |627 |Disease |0.55843806|
|vitamin D deficiency|633 |652 |Disease |0.5240673 |
|weekly |668 |673 |Frequency |0.66599834|
|supplement |675 |684 |Drug |0.5972447 |
|vitamin D |689 |697 |Drug |0.86098486|
|1000 mcg |703 |710 |Dosage |0.9711681 |
|b12 |712 |714 |Drug |0.83438593|
|daily |716 |720 |Frequency |0.93205667|
|homeopathy medicine |734 |752 |Drug |0.5354072 |
|40 days |758 |764 |Duration |0.9758632 |
|after 30 days |784 |796 |DateTime |0.8755744 |
|TSH |802 |804 |Test |0.98604935|
|now |813 |815 |DateTime |0.8677737 |
|weakness |850 |857 |Symptom |0.9838275 |
|depression |863 |872 |PsychologicalCondition|0.89689744|
|last week |913 |921 |DateTime |0.8067833 |
|thyroid |1075 |1081|BodyPart |0.8369291 |
+--------------------+-----+----+----------------------+----------+
Model Information
| Model Name: | zeroshot_ner_vop_large |
| Compatibility: | Healthcare NLP 5.5.1+ |
| License: | Licensed |
| Edition: | Official |
| Language: | en |
| Size: | 1.6 GB |
Benchmarking
label precision recall f1-score support
AdmissionDischarge 0.4507 0.9697 0.6154 33
Age 0.8991 0.9635 0.9302 657
Allergen 0.3717 0.9130 0.5283 46
BodyPart 0.8916 0.9583 0.9238 2953
ClinicalDept 0.7534 0.9424 0.8373 295
DateTime 0.8348 0.8969 0.8648 4395
Disease 0.8917 0.7993 0.8430 2277
Dosage 0.6617 0.8898 0.7590 499
Drug 0.8271 0.9246 0.8732 1340
Duration 0.7676 0.8581 0.8103 2255
Employment 0.9338 0.9522 0.9429 1214
Form 0.6627 0.9621 0.7848 290
Frequency 0.6645 0.9149 0.7698 1093
Gender 0.9564 0.9675 0.9619 1293
Laterality 0.6961 0.9548 0.8052 619
MedicalDevice 0.5978 0.6781 0.6354 320
Modifier 0.3887 0.9112 0.5449 1092
O 0.9849 0.9468 0.9655 114576
Procedure 0.6593 0.7295 0.6926 573
PsychologicalCondition 0.8029 0.9236 0.8590 419
RaceEthnicity 0.8810 1.0000 0.9367 37
Substance 0.8292 0.9618 0.8906 419
Symptom 0.7636 0.8518 0.8053 4501
Test 0.8112 0.8481 0.8292 1560
Treatment 0.3214 0.6562 0.5315 192
Vaccine 0.9444 0.7391 0.8293 46
accuracy - - 0.9352 142994
macro avg 0.7403 0.8890 0.7950 142994
weighted avg 0.9471 0.9352 0.9393 142994