Few Shot Age Group Classification

Description

Few-shot classification in NLP refers to the task of building text classification models that can learn from only a small number of labeled examples per class. It aims to mimic the human ability to learn concepts from just a few examples.

The original dataset included thousands of labelled text; for the training of this Few Shot Classifier model, only a small portion of the original dataset was used and the benchmarks are satisfactory.

This Few Shot Classifier model is trained for analyzing the age group of a person mentioned in health documents. Age of the person may or may not be mentioned explicitly in the training dataset.

The Text Classifier model has been trained using in-house annotated health-related text that have been labeled with three different classes:

Adult: A person who is fully grown or developed. Typically refers to someone who is 18 years or older,

Child: Requires intervention, urgent, not life-threatening cases.

Unknown: Not possible to comprehend/figure out the age group from the given text.

Predicted Entities

Adult, Child, Unknown

Copy S3 URI

How to use

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

bert_sent = BertSentenceEmbeddings.pretrained("sbiobert_base_cased_mli", "en", "clinical/models")\
    .setInputCols(["document"])\
    .setOutputCol("sentence_embeddings")

few_shot_classifier = FewShotClassifierModel.pretrained("few_shot_classifier_age_group_sbiobert_cased_mli", "en", "clinical/models")\
    .setInputCols(["sentence_embeddings"])\
    .setOutputCol("prediction")

clf_Pipeline = Pipeline(stages=[
    document_assembler,
    bert_sent,
    few_shot_classifier
])

data = spark.createDataFrame([["""A patient presented with complaints of chest pain and shortness of breath. The medical history revealed the patient had a smoking habit for over 30 years, and was diagnosed with hypertension two years ago. After a detailed physical examination, the doctor found a noticeable wheeze on lung auscultation and prescribed a spirometry test, which showed irreversible airway obstruction. The patient was diagnosed with Chronic obstructive pulmonary disease (COPD) caused by smoking."""],
 ["""Hi, wondering if anyone has had a similar situation. My 1 year old daughter has the following; loose stools/ pale stools, elevated liver enzymes, low iron.  5 months and still no answers from drs. """],
 ["""Hi have chronic gastritis from 4 month(confirmed by endoscopy).I do not have acid reflux.Only dull ache above abdomen and left side of chest.I am on reberprozole and librax.My question is whether chronic gastritis is curable or is it a lifetime condition?I am loosing hope because this dull ache is not going away.Please please reply"""]]).toDF("text")

result = clf_Pipeline.fit(data).transform(data)
val document_assembler =new DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")

val sentence_embeddings = new BertSentenceEmbeddings.pretrained("sbiobert_base_cased_mli", "en", "clinical/models")
    .setInputCols("document")
    .setOutputCol("sentence_embeddings")

val few_shot_classifier = new FewShotClassifierModel.pretrained("few_shot_classifier_age_group_sbiobert_cased_mli", "en", "clinical/models")
    .setInputCols("sentence_embeddings")
    .setOutputCol("prediction")

val clf_Pipeline = new Pipeline().setStages(Array(document_assembler, sentence_embeddings, few_shot_classifier))

val data = Seq(Array("A patient presented with complaints of chest pain and shortness of breath. The medical history revealed the patient had a smoking habit for over 30 years, and was diagnosed with hypertension two years ago. After a detailed physical examination, the doctor found a noticeable wheeze on lung auscultation and prescribed a spirometry test, which showed irreversible airway obstruction. The patient was diagnosed with Chronic obstructive pulmonary disease (COPD) caused by smoking.", "Hi, wondering if anyone has had a similar situation. My 1 year old daughter has the following; loose stools/ pale stools, elevated liver enzymes, low iron.  5 months and still no answers from drs.", "Hi have chronic gastritis from 4 month(confirmed by endoscopy).I do not have acid reflux.Only dull ache above abdomen and left side of chest.I am on reberprozole and librax.My question is whether chronic gastritis is curable or is it a lifetime condition?I am loosing hope because this dull ache is not going away.Please please reply")).toDS().toDF("text")

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

Results

+------------------------------------------------------------------------------------------------------------------------------------------------------+---------+
|                                                                                                                                                  text|   result|
+------------------------------------------------------------------------------------------------------------------------------------------------------+---------+
|A patient presented with complaints of chest pain and shortness of breath. The medical history revealed the patient had a smoking habit for over 30...|  [Adult]|
|Hi, wondering if anyone has had a similar situation. My 1 year old daughter has the following; loose stools/ pale stools, elevated liver enzymes, l...|  [Child]|
|Hi have chronic gastritis from 4 month(confirmed by endoscopy).I do not have acid reflux.Only dull ache above abdomen and left side of chest.I am o...|[Unknown]|
+------------------------------------------------------------------------------------------------------------------------------------------------------+---------+

Model Information

Model Name: few_shot_classifier_age_group_sbiobert_cased_mli
Compatibility: Healthcare NLP 5.0.1+
License: Licensed
Edition: Official
Input Labels: [sentence_embeddings]
Output Labels: [prediction]
Language: en
Size: 47.2 KB

References

In-house annotated health-related text.

Benchmarking

       label  precision    recall  f1-score   support
       Adult       0.76      0.72      0.74       329
       Child       0.87      0.91      0.89       194
     Unknown       0.74      0.77      0.76       325
    accuracy       -         -         0.78       848
   macro-avg       0.79      0.80      0.80       848
weighted-avg       0.78      0.78      0.78       848