Also Known as SkimLit or PubMed.
In this notebook, I replicated a state-of-the-art (NLP) model to breakdown research abstracts into smaller, more understandable pieces to help researchers skim through literatures.
Data Source:
Model Score on validataion data:
- Accuracy on previously unseen data of %86.27
- precision on previously unseen data of 0.863
- recall on previously unseen data of 0.863
- F1 on previously unseen data of 0.860
Problem Type:
- Text Classification