Description
The legclf_exchange_agreement_bert model is a Bert Sentence Embeddings Document Classifier used to classify if the document belongs to the class exchange-agreement or not (Binary Classification).
Unlike the Longformer model, this model is lighter in terms of inference time.
Predicted Entities
exchange-agreement, other
How to use
document_assembler = nlp.DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
embeddings = nlp.BertSentenceEmbeddings.pretrained("sent_bert_base_cased", "en")\
.setInputCols("document")\
.setOutputCol("sentence_embeddings")
doc_classifier = legal.ClassifierDLModel.pretrained("legclf_exchange_agreement_bert", "en", "legal/models")\
.setInputCols(["sentence_embeddings"])\
.setOutputCol("category")
nlpPipeline = nlp.Pipeline(stages=[
document_assembler,
embeddings,
doc_classifier])
df = spark.createDataFrame([["YOUR TEXT HERE"]]).toDF("text")
model = nlpPipeline.fit(df)
result = model.transform(df)
Results
+-------+
|result|
+-------+
|[exchange-agreement]|
|[other]|
|[other]|
|[exchange-agreement]|
Model Information
| Model Name: | legclf_exchange_agreement_bert |
| Compatibility: | Legal NLP 1.0.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence_embeddings] |
| Output Labels: | [class] |
| Language: | en |
| Size: | 22.8 MB |
References
Legal documents, scrapped from the Internet, and classified in-house + SEC documents
Benchmarking
label precision recall f1-score support
exchange-agreement 0.97 0.75 0.85 40
other 0.89 0.99 0.94 82
accuracy - - 0.91 122
macro-avg 0.93 0.87 0.89 122
weighted-avg 0.92 0.91 0.91 122