Table 4 Confusion matrix for test results for the PubMed 200k dataset, yielded by the fastText model with sentence context and numeric sentence position
From: Fast and scalable neural embedding models for biomedical sentence classification
True label | Predicted label | ||||
|---|---|---|---|---|---|
| Â | Objective | Background | Methods | Results | Conclusions |
Objective | 1704 (72%) | 573 | 98 | 0 | 2 |
Background | 471 | 2051 (77%) | 91 | 8 | 42 |
Methods | 25 | 36 | 9375 (96%) | 296 | 18 |
Results | 3 | 3 | 395 | 9744 (95%) | 131 |
Conclusions | 1 | 23 | 13 | 203 | 4186 (95%) |