Description
This model is a PHS-BERT based text classification model that can classify depression level of social media text into three levels: no-depression, minimum, high-depression.
Predicted Entities
no-depression, minimum, high-depression
How to use
document_assembler = DocumentAssembler() \
.setInputCol("text") \
.setOutputCol("document")
tokenizer = Tokenizer() \
.setInputCols(["document"]) \
.setOutputCol("token")
sequenceClassifier = MedicalBertForSequenceClassification.pretrained("bert_sequence_classifier_depression", "en", "clinical/models")\
.setInputCols(["document","token"])\
.setOutputCol("class")
pipeline = Pipeline(stages=[
document_assembler,
tokenizer,
sequenceClassifier
])
data = spark.createDataFrame([
["None that I know of. Any mental health issue needs to be cared for like any other health issue. Doctors and medications can help."],
["I don’t know. Was this okay? Should I hate him? Or was it just something new? I really don’t know what to make of the situation."],
["It makes me so disappointed in myself because I hate what I've become and I hate feeling so helpless."]
]).toDF("text")
result = pipeline.fit(data).transform(data)
result.select("text", "class.result").show(truncate=False)
val documenter = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val tokenizer = new Tokenizer()
.setInputCols(Array("document"))
.setOutputCol("token")
val sequenceClassifier = MedicalBertForSequenceClassification.pretrained("bert_sequence_classifier_depression", "en", "clinical/models")
.setInputCols(Array("document","token"))
.setOutputCol("class")
val pipeline = new Pipeline().setStages(Array(documenter, tokenizer, sequenceClassifier))
val data = Seq(Array(
"None that I know of. Any mental health issue needs to be cared for like any other health issue. Doctors and medications can help.",
"I don’t know. Was this okay? Should I hate him? Or was it just something new? I really don’t know what to make of the situation.",
"It makes me so disappointed in myself because I hate what I've become and I hate feeling so helpless."
)).toDS.toDF("text")
val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("en.classify .bert_sequence.depression").predict("""None that I know of. Any mental health issue needs to be cared for like any other health issue. Doctors and medications can help.""")
Results
+---------------------------------------------------------------------------------------------------------------------------------+-----------------+
|text |result |
+---------------------------------------------------------------------------------------------------------------------------------+-----------------+
|None that I know of. Any mental health issue needs to be cared for like any other health issue. Doctors and medications can help.|[no-depression] |
|I don’t know. Was this okay? Should I hate him? Or was it just something new? I really don’t know what to make of the situation. |[minimum] |
|It makes me so disappointed in myself because I hate what I've become and I hate feeling so helpless. |[high-depression]|
+---------------------------------------------------------------------------------------------------------------------------------+-----------------+
Model Information
| Model Name: | bert_sequence_classifier_depression |
| Compatibility: | Healthcare NLP 4.0.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [document, token] |
| Output Labels: | [class] |
| Language: | en |
| Size: | 1.3 GB |
| Case sensitive: | true |
| Max sentence length: | 128 |
Benchmarking
label precision recall f1-score support
no-depression 0.99 0.99 0.99 98
minimum 0.85 0.86 0.85 155
high-depression 0.81 0.80 0.81 119
accuracy - - 0.87 372
macro-avg 0.88 0.88 0.88 372
weighted-avg 0.87 0.87 0.87 372