Final Notebooks are in the folder https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/
The folder contains:
1 - Four LSTM models notebooks for mention classfication. The models uses different BERT models for feature extraction
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-classification-biobert.ipynb
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-classification-umlsbert.ipynb
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-classification-scibert.ipynb
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-classification-coder.ipynb
The four models are using the identical methodologies except the use of the BERT model
2 - Two LSTM models notebooks for mention detection. The models uses different BERT models for feature extraction
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-detect-scibert.ipynb
- https://github.com/mhmdrdwn/thesis/blob/main/notebooks/final_work/mention-detection-coder.ipynb
The two models are using the identical methodologies except the use of the BERT model
3 - Nearest Neighbour Search
4 - Visualization of the results of all the models
Some of the methods here are following and inspired after the following studies:
- D. Loureiro, A. Mário Jorge, MedLinker: Medical Entity Linking with Neural Representations and Dictionary Matching, 2020, https://github.com/danlou/MedLinker.
- K. C. Fraser, I. Nejadgholi, B. De Bruijn, M. Li, A. LaPlante, K. Zine El Abidine, Extracting UMLS Concepts from Medical Text Using General and Domain-Specific Deep Learning Models, 2019, https://arxiv.org/abs/1910.01274
Project based on the cookiecutter data science project template. #cookiecutterdatascience