Description
This is a Sentiment Analysis model which retrieves 3 sentiments (positive, negative or neutral) from Auditors’ comments.
Predicted Entities
positive, negative, neutral
How to use
documentAssembler = nlp.DocumentAssembler() \
.setInputCol("sentence") \
.setOutputCol("document")
embeddings = nlp.BertSentenceEmbeddings.pretrained("sent_bert_base_cased", "en") \
.setInputCols("document") \
.setOutputCol("sentence_embeddings")
sentiment = nlp.ClassifierDLModel.pretrained("finclf_auditor_sentiment_analysis", "en", "finance/models") \
.setInputCols("sentence_embeddings") \
.setOutputCol("category")
pipeline = nlp.Pipeline() \
.setStages(
[
documentAssembler,
embeddings,
sentiment
]
)
pipelineModel = pipeline.fit(sdf_test)
res = pipelineModel.transform(sdf_test)
res.select('sentence', 'category.result').show(truncate=100)
Results
+----------------------------------------------------------------------------------------------------+----------+
| sentence| result|
+----------------------------------------------------------------------------------------------------+----------+
|In our opinion, the consolidated financial statements referred to above present fairly..............|[positive]|
+----------------------------------------------------------------------------------------------------+----------+
Model Information
| Model Name: | finclf_auditor_sentiment_analysis |
| Compatibility: | Finance NLP 1.0.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence_embeddings] |
| Output Labels: | [category] |
| Language: | en |
| Size: | 23.1 MB |
References
Propietary auditors’ reports
Benchmarking
label precision recall f1-score support
negative 0.66 0.78 0.72 124
neutral 0.88 0.77 0.82 559
positive 0.65 0.76 0.70 286
accuracy - - 0.77 969
macro-avg 0.73 0.77 0.74 969
weighted-avg 0.78 0.77 0.77 969