Description
NER model that detects professions and occupations in Spanish texts. Trained with the embeddings_scielowiki_300d embeddings, and the same WordEmbeddingsModel is needed in the pipeline.
Predicted Entities
ACTIVIDAD, PROFESION, SITUACION_LABORAL
Live Demo Open in Colab Download
How to use
document_assembler = DocumentAssembler()\
.setInputCol('text')\
.setOutputCol('document')
sentence = SentenceDetector() \
.setInputCols(["document"]) \
.setOutputCol("sentence")
tokenizer = Tokenizer() \
.setInputCols(["sentence"]) \
.setOutputCol("token")
embedings_stage = WordEmbeddingsModel.pretrained("embeddings_scielowiki_300d", "es", "clinical/models")\
.setInputCols(["document","token"])\
.setOutputCol("word_embeddings")
clinical_ner = MedicalNerModel.pretrained("meddroprof_scielowiki", "es", "clinical/models")\
.setInputCols(["sentence", "token", "embeddings"])\
.setOutputCol("ner")
ner_converter = NerConverter() \
.setInputCols(['document', 'token', 'ner']) \
.setOutputCol('ner_chunk')
pipeline = Pipeline(stages=[
document_assembler,
sentence,
tokenizer,
embedings_stage,
clinical_ner,
ner_converter
])
example = spark.createDataFrame(pd.DataFrame({'text': ["""La paciente es la mayor de 2 hermanos, tiene un hermano de 13 años estudiando 1o ESO. Sus padres son ambos ATS , trabajan en diferentes centros de salud estudiando 1o ESO"""]}))
result = pipeline.fit(example).transform(example)
val document_assembler = new DocumentAssembler()
.setInputCol('text')
.setOutputCol('document')
val sentence = new SentenceDetector()
.setInputCols(["document"])
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols(["sentence"])
.setOutputCol("token")
val embedings_stage = WordEmbeddingsModel.pretrained("embeddings_scielowiki_300d", "es", "clinical/models")
.setInputCols(["document","token"])
.setOutputCol("word_embeddings")
val clinical_ner = MedicalNerModel.pretrained("meddroprof_scielowiki", "es", "clinical/models")
.setInputCols(["sentence", "token", "embeddings"])
.setOutputCol("ner")
val ner_converter = new NerConverter()
.setInputCols(['document', 'token', 'ner'])
.setOutputCol('ner_chunk')
val pipeline = new Pipeline().setStages(Array(document_assembler, sentence, tokenizer, embedings_stage, clinical_ner, ner_converter))
val data = Seq("""La paciente es la mayor de 2 hermanos, tiene un hermano de 13 años estudiando 1o ESO. Sus padres son ambos ATS , trabajan en diferentes centros de salud estudiando 1o ESO""").toDS.toDF("text")
val result = pipeline.fit(data).transform(data)
Results
+--------------------+----------+-------------------+
| text| token| prediction|
+--------------------+----------+-------------------+
|La paciente es la...| La| O|
|La paciente es la...| paciente| O|
|La paciente es la...| es| O|
|La paciente es la...| la| O|
|La paciente es la...| mayor| O|
|La paciente es la...| de| O|
|La paciente es la...| 2| O|
|La paciente es la...| hermanos| O|
|La paciente es la...| ,| O|
|La paciente es la...| tiene| O|
|La paciente es la...| un| O|
|La paciente es la...| hermano| O|
|La paciente es la...| de| O|
|La paciente es la...| 13| O|
|La paciente es la...| años| O|
|La paciente es la...|estudiando|B-SITUACION_LABORAL|
|La paciente es la...| 1o|I-SITUACION_LABORAL|
|La paciente es la...| ESO|I-SITUACION_LABORAL|
|La paciente es la...| .| O|
|La paciente es la...| Sus| O|
|La paciente es la...| padres| O|
|La paciente es la...| son| O|
|La paciente es la...| ambos| O|
|La paciente es la...| ATS| B-PROFESION|
|La paciente es la...| ,| O|
|La paciente es la...| trabajan| B-PROFESION|
|La paciente es la...| en| I-PROFESION|
|La paciente es la...|diferentes| I-PROFESION|
|La paciente es la...| centros| I-PROFESION|
|La paciente es la...| de| I-PROFESION|
|La paciente es la...| salud| I-PROFESION|
|La paciente es la...|estudiando|B-SITUACION_LABORAL|
|La paciente es la...| 1o|I-SITUACION_LABORAL|
|La paciente es la...| ESO|I-SITUACION_LABORAL|
+--------------------+----------+-------------------+
Model Information
| Model Name: | meddroprof_scielowiki |
| Compatibility: | Spark NLP for Healthcare 3.1.3+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, word_embeddings] |
| Output Labels: | [ner] |
| Language: | es |
| Dependencies: | embeddings_scielowiki_300d |
Data Source
The model was trained with the MEDDOPROF data set:
The MEDDOPROF corpus is a collection of 1844 clinical cases from over 20 different specialties annotated with professions and employment statuses. The corpus was annotated by a team composed of linguists and clinical experts following specially prepared annotation guidelines, after several cycles of quality control and annotation consistency analysis before annotating the entire dataset. Figure 1 shows a screenshot of a sample manual annotation generated using the brat annotation tool.
Reference:
@article{meddoprof,
title={NLP applied to occupational health: MEDDOPROF shared task at IberLEF 2021 on automatic recognition, classification and normalization of professions and occupations from medical texts},
author={Lima-López, Salvador and Farré-Maduell, Eulàlia and Miranda-Escalada, Antonio and Brivá-Iglesias, Vicent and Krallinger, Martin},
journal = {Procesamiento del Lenguaje Natural},
volume = {67},
year={2021}
}
Benchmarking
label precision recall f1-score support
B-ACTIVIDAD 0.82 0.36 0.50 25
B-PROFESION 0.87 0.75 0.81 634
B-SITUACION_LABORAL 0.79 0.67 0.72 310
I-ACTIVIDAD 0.86 0.43 0.57 58
I-PROFESION 0.87 0.80 0.83 944
I-SITUACION_LABORAL 0.74 0.71 0.73 407
O 1.00 1.00 1.00 139880
accuracy - - 0.99 142258
macro-avg 0.85 0.67 0.74 142258
weighted-avg 0.99 0.99 0.99 142258