Table 4 Results for Extended NeuroNER entity classification using combinations of embeddings models on PharmaCoNER test dataset
Experiment | Embedding model | Precision (%) | Recall (%) | F-score (%) |
|---|---|---|---|---|
Run 4 | SNOMED-SBC + Reddit | 83.52 | 74.97 | 79.02 |
Run 2 | W2V-SBWC + Reddit | 83.85 | 75.75 | 79.60 |
Run 3 | FastText-SBWC + Reddit | 84.70 | 77.31 | 80.84 |
Run 1 | FastText-SBC + Reddit | 89.13 | 82.61 | 85.75 |
Out of task | Scielo+Wiki cased + Reddit | 86.69 | 82.72 | 84.66 |
Out of task | PubMed and PMC + Reddit | 87.23 | 76.98 | 81.79 |
Out of task | FastText 2M + Reddit | 84.04 | 77.55 | 80.67 |