Edit on GitHub
  • John Snow Labs
  • Mar 27, 2023

Financial Target NER (Codalab)

  • bert
  • finance
  • ner
  • es
  • licensed
  • tensorflow

Description

This Spanish NER model will identify the label TARGET from a financial statement. This model is trained from the competition - IBERLEF 2023 Task - FinancES. Financial Targeted Sentiment Analysis in Spanish. We have used the participation dataset which is a small subset of the main one to train this model.

Predicted Entities

TARGET

Copy S3 URI

How to use

documentAssembler = nlp.DocumentAssembler()\
  .setInputCol("text")\
  .setOutputCol("document")

tokenizer = nlp.Tokenizer()\
  .setInputCols("document")\
  .setOutputCol("token")
  
tokenClassifier = finance.BertForTokenClassification.pretrained("finner_bert_target","es","finance/models")\
  .setInputCols("token", "document")\
  .setOutputCol("label")\
  .setCaseSensitive(True)

converter = finance.NerConverterInternal()\
    .setInputCols(["document", "token", "label"])\
    .setOutputCol("ner")

pipeline =  nlp.Pipeline(
    stages=[
  documentAssembler,
  tokenizer,
  tokenClassifier,
  converter
    ]
)

Results

+-----------+------+
|chunk      |entity|
+-----------+------+
|Presupuesto|TARGET|
|populista  |TARGET|
+-----------+------+

Model Information

Model Name: finner_bert_target
Compatibility: Finance NLP 1.0.0+
License: Licensed
Edition: Official
Input Labels: [sentence, token]
Output Labels: [ner]
Language: es
Size: 406.6 MB
Case sensitive: true
Max sentence length: 128

References

https://codalab.lisn.upsaclay.fr/competitions/10052#learn_the_details

Benchmarking

 
labels              precision    recall  f1-score   support
    B-TARGET       0.76      0.82      0.79       435
   micro-avg       0.76      0.82      0.79       435
   macro-avg       0.76      0.82      0.79       435
weighted-avg       0.76      0.82      0.79       435