Employing GNNs for multi-label classification of biomedical articles
Welcome to the documentation of this project.
This project focuses on the multi-label classification of academic articles using GNNs and label-guided embeddings aimed at mitigating the performance decrease resulting from strongly imbalanced labels.
The documentation will cover the different aspects of the project, namely:
- Setting up the project
- Creating the right settings
- Preparing the data
- Training a model
- Optimising a model
General Remarks
This project was created as a 3-month Master's Thesis project for the program Information Studies - Data Science. The project was done in collaboration with Elsevier under the supervision of Bruno Martins, Chieling Yueh and Prof. dr. Paul Groth.
Contact
If you have any questions or comments regarding this project, contact me through:
- Email: rvdb7345@gmail.com