Evaluation

Evaluation will be done by comparing the automatically generated results to the results generated by manual annotation of experts, with the main metrics being micro-averaged precision, recall and F1-scores.

The MultiCardioNER evaluation library is available on GitHub.