PE NLP pipeline for training and testing. it contains functions including:
- Save annotation dataframes to annotation files
- Read ann file of selected document
- Convert annotations to dataframe
- Convert dataframe to Annotation instance
- Mention level evaluation
D-dimer NLP pipeline for training and testing. it contains similar functions as PE pipeline.
Combined PE and d-dimer pipeline, and save the annotation results to output.csv.
Covert csv file format to PyContextNLP modifier_rules format.
Modified DocumentClassifier.py.
A wrapper of PyContextNLP using PyRush as the sentence segmenter. It return entity annotations, relationships and document classification.
This module is used for evaluation.
New PE NLP pipeline for training, testing and deployment. Output file: out_table.csv.
New d-dimer pipeline for training, testing and deployment. Output file: out_table_dimer.csv.