### Comparative analysis of neural methods for entity normalization in the biological field.
### Comparative analysis of NLP neural methods for the entity normalization task in the biological field.
This work is an attempt to make interoperable multiple datasets and methods in the entity normalization task, in order to evaluate the robustness of those methods.
This work is an attempt to make interoperable multiple datasets and methods in a NLP Sequence Classification subtask: the entity normalization task, in order to evaluate the robustness of those methods.
This repository is made to store and make accessible all scripts and resulting data.
Two SoTA normalization methods (2022) have been adapted so far:
Two State-of-The-Art normalization methods (2022) have been adapted so far:
-[Lightweight](https://github.com/tigerchen52/Biomedical-Entity-Linking)[1], a method using Word2Vec embeddings from the paper [A Lightweight Neural Model for Biomedical Entity Linking](https://arxiv.org/abs/2012.08844)
-[BioSyn](https://github.com/dmis-lab/BioSyn)[2], a method using BioBERT embeddings and morpho-syntaxic representation of mentions and labels from the paper [Biomedical Entity Representations with Synonym Marginalization](https://arxiv.org/abs/2005.00239)
