Description
This is a Multilabel Document Classification model, which can be used to identify up to 15 classes in texts. The classes are the following:
- terminations
- assigns
- notices
- amendments
- waivers
- survival
- successors
- governing laws
- severability
- expenses
- assignments
- warranties
- representations
- entire agreements
- counterparts
Predicted Entities
terminations
How to use
document = nlp.DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
embeddings = nlp.E5Embeddings.pretrained("legembedding_e5_base", "en", "legal/models")\
.setInputCols(["document"])\
.setOutputCol("sentence_embeddings")
multiClassifier = nlp.MultiClassifierDLModel.pretrained("legmulticlf_edgar_le", "en", "legal/models") \
.setInputCols(["document", "sentence_embeddings"]) \
.setOutputCol("class")
ledgar_pipeline = nlp.Pipeline(
stages=[document,
embeddings,
multiClassifier])
light_pipeline = LightPipeline(ledgar_pipeline.fit(spark.createDataFrame([['']]).toDF("text")))
result = light_pipeline.annotate("""(a) No failure or delay by the Administrative Agent or any Lender in exercising any right or power hereunder shall operate as a waiver thereof, nor shall any single or partial exercise of any such right or power, or any abandonment or discontinuance of steps to enforce such a right or power, preclude any other or further exercise thereof or the exercise of any other right or power. The rights and remedies of the Administrative Agent and the Lenders hereunder are cumulative and are not exclusive of any rights or remedies that they would otherwise have. No waiver of any provision of this Agreement or consent to any departure by the Borrower therefrom shall in any event be effective unless the same shall be permitted by paragraph (b) of this Section, and then such waiver or consent shall be effective only in the specific instance and for the purpose for which given. Without limiting the generality of the foregoing, the making of a Loan shall not be construed as a waiver of any Default, regardless of whether the Administrative Agent or any Lender may have had notice or knowledge of such Default at the time.""")
result["class"]
Results
['waivers', 'amendments']
Model Information
| Model Name: | legmulticlf_edgar_le |
| Compatibility: | Legal NLP 1.0.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence_embeddings] |
| Output Labels: | [class] |
| Language: | en |
| Size: | 14.0 MB |
References
Ledgar dataset, available at https://metatext.io/datasets/ledgar, with in-house data
Benchmarking
Classification report:
precision recall f1-score support
0 0.89 0.89 0.89 1066
1 0.83 0.65 0.73 333
2 0.80 0.81 0.80 537
3 0.99 0.99 0.99 918
4 0.98 0.98 0.98 1049
5 0.99 0.97 0.98 339
6 1.00 0.99 0.99 1274
7 0.98 0.98 0.98 926
8 0.91 0.92 0.91 437
9 0.98 0.97 0.98 922
10 0.89 0.88 0.88 674
11 0.95 0.96 0.95 566
12 0.92 0.79 0.85 354
13 0.89 0.87 0.88 725
14 0.88 0.78 0.83 365
micro avg 0.94 0.92 0.93 10485
macro avg 0.93 0.89 0.91 10485
weighted avg 0.94 0.92 0.93 10485
samples avg 0.93 0.94 0.93 10485