Description
This is a md version of Indian News Sentiment Analysis Text Classifier, which will retrieve if a text is either expression a Positive Emotion or a Negative one.
Predicted Entities
POSITIVE, NEGATIVE
How to use
document_assembler = nlp.DocumentAssembler() \
.setInputCol("text") \
.setOutputCol("document")
tokenizer = nlp.Tokenizer() \
.setInputCols(["document"]) \
.setOutputCol("token")
classifierdl = finance.BertForSequenceClassification.pretrained("finclf_indian_news_sentiment_medium","en", "finance/models")\
.setInputCols(["document", "token"])\
.setOutputCol("label")
bert_clf_pipeline = nlp.Pipeline(stages=[document_assembler,
tokenizer,
classifierdl])
text = ["Eliminating shadow economy to have positive impact on GDP : Arun Jaitley"]
empty_df = spark.createDataFrame([[""]]).toDF("text")
model = bert_clf_pipeline.fit(empty_df)
res = model.transform(spark.createDataFrame([text]).toDF("text"))
Results
+------------------------------------------------------------------------+----------+
|text |result |
+------------------------------------------------------------------------+----------+
|Eliminating shadow economy to have positive impact on GDP : Arun Jaitley|[POSITIVE]|
+------------------------------------------------------------------------+----------+
Model Information
| Model Name: | finclf_indian_news_sentiment_medium |
| Compatibility: | Finance NLP 1.0.0+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [document, token] |
| Output Labels: | [class] |
| Language: | en |
| Size: | 412.3 MB |
| Case sensitive: | true |
| Max sentence length: | 128 |
References
An in-house augmented version of this dataset
Benchmarking
label precision recall f1-score support
NEGATIVE 0.85 0.86 0.86 10848
POSITIVE 0.83 0.83 0.83 9202
accuracy - - 0.84 20050
macro-avg 0.84 0.84 0.84 20050
weighted-avg 0.84 0.84 0.84 20050