Generate-and-Rank-ConNorm

Generate-and-Rank Framework with Semantic Type Regularization for Biomedical Concept Normalization

Setup

• We recommend Python 3.7 or higher. The code does not work with Python 2.7.
• BERT-based multi-class classifier and BERT-based listwise classifier are implemented with PyTorch (at least 1.3.0) using transformers v3.0.2.
• Lucene-based dictionary look-up system is developed using the lucene-8.1.1.

Getting Started

Candidate Generator

• BERT-based multi-class classifier: Generator/Multiclass/
• Lucene-based dictionary look-up system: Generator/Lucene/

Candidate ranker

• BERT-based listwise classifier: Ranker/