ALBERT Token Classification Large - NER CoNLL (albert_large_token_classifier_conll03)

Description

ALBERT Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks.

albert_large_token_classifier_conll03 is a fine-tuned ALBERT model that is ready to use for Named Entity Recognition and achieves state-of-the-art performance for the NER task. This model has been trained to recognize four types of entities: location (LOC), organizations (ORG), person (PER), and Miscellaneous (MISC).

We used TFAlbertForTokenClassification to train this model and used AlbertForTokenClassification annotator in Spark NLP 🚀 for prediction at scale!

Predicted Entities

LOC, ORG, PER, MISC

Download Copy S3 URI

How to use

document_assembler = DocumentAssembler() \
.setInputCol('text') \
.setOutputCol('document')

tokenizer = Tokenizer() \
.setInputCols(['document']) \
.setOutputCol('token')

tokenClassifier = AlbertForTokenClassification \
.pretrained('albert_large_token_classifier_conll03', 'en') \
.setInputCols(['token', 'document']) \
.setOutputCol('ner') \
.setCaseSensitive(True) \
.setMaxSentenceLength(512)

# since output column is IOB/IOB2 style, NerConverter can extract entities
ner_converter = NerConverter() \
.setInputCols(['document', 'token', 'ner']) \
.setOutputCol('entities')

pipeline = Pipeline(stages=[
document_assembler, 
tokenizer,
tokenClassifier,
ner_converter
])

example = spark.createDataFrame([['My name is John!']]).toDF("text")
result = pipeline.fit(example).transform(example)
val document_assembler = DocumentAssembler() 
.setInputCol("text") 
.setOutputCol("document")

val tokenizer = Tokenizer() 
.setInputCols("document") 
.setOutputCol("token")

val tokenClassifier = AlbertForTokenClassification.pretrained("albert_large_token_classifier_conll03", "en")
.setInputCols("document", "token")
.setOutputCol("ner")
.setCaseSensitive(true)
.setMaxSentenceLength(512)

// since output column is IOB/IOB2 style, NerConverter can extract entities
val ner_converter = NerConverter() 
.setInputCols("document", "token", "ner") 
.setOutputCol("entities")

val pipeline = new Pipeline().setStages(Array(document_assembler, tokenizer, tokenClassifier, ner_converter))

val example = Seq.empty["My name is John!"].toDS.toDF("text")

val result = pipeline.fit(example).transform(example)
import nlu
nlu.load("en.classify.token_albert_large_token_classifier_conll03").predict("""My name is John!""")

Results

+------------------------------------------------------------------------------------+
|result                                                                              |
+------------------------------------------------------------------------------------+
|[B-PER, I-PER, O, O, O, B-LOC, O, O, O, B-LOC, O, O, O, O, B-PER, O, O, O, O, B-LOC]|
+------------------------------------------------------------------------------------+

Model Information

Model Name: albert_large_token_classifier_conll03
Compatibility: Spark NLP 3.3.0+
License: Open Source
Edition: Official
Input Labels: [token, document]
Output Labels: [ner]
Language: en
Case sensitive: false
Max sentense length: 512

Data Source

https://www.clips.uantwerpen.be/conll2003/ner/

Benchmarking

precision    recall  f1-score   support

B-LOC       0.97      0.96      0.96      1837
B-MISC       0.89      0.90      0.89       922
B-ORG       0.89      0.93      0.91      1341
B-PER       0.91      0.97      0.94      1842
I-LOC       0.94      0.91      0.92       257
I-MISC       0.85      0.79      0.82       346
I-ORG       0.88      0.88      0.88       751
I-PER       0.98      0.90      0.94      1307
O       0.99      0.99      0.99     42759

accuracy                           0.98     51362
macro avg       0.92      0.91      0.92     51362
weighted avg       0.98      0.98      0.98     51362



processed 51362 tokens with 5942 phrases; found: 6158 phrases; correct: 5426.
accuracy:  92.57%; (non-O)
accuracy:  98.25%; precision:  88.11%; recall:  91.32%; FB1:  89.69
LOC: precision:  95.82%; recall:  95.97%; FB1:  95.89  1840
MISC: precision:  84.96%; recall:  86.98%; FB1:  85.96  944
ORG: precision:  84.40%; recall:  89.19%; FB1:  86.73  1417
PER: precision:  85.08%; recall:  90.39%; FB1:  87.65  1957