CORD-NER methods are domain-independent that can be applied to corpus in different domains. Regarding "De-identification and Heart Disease Risk Factors", we evaluated NER performance comparison between SciSpacy and our annotation results on the N2C2 Dataset (with 7-9% improvements over previous approaches) and visualized the results on TensorBoard. CORD-NER annotation is a combination from 4 sources: (Reference: here )

1. Pretrained NER on 18 General Entity Types: Spacy
2. Pretrained NER on 18 Biomedical Entity Types: SciSpacy
3. Knowledge Base (KB)-Guided NER on 127 Biomedical Entity Types
4. Seed-Guided NER on 9 New Entity Types

NeuroNER leverages the state-of-the-art prediction capabilities of deep learning and enables the users to create or modify annotations for a new or existing corpus. The NeuroNER engine is based on artificial neural networks (ANNs). Specifically, it relies on a variant of recurrent neural network (RNN) called long short-term memory (LSTM). The NER engine's ANN contains three layers: (Reference: here )

1. Character-enhanced token-embedding layer
2. Label prediction layer
3. Label sequence optimization layer

On the N2C2 Dataset (Obesity Challenge Factors), we have trained the neural network that performs the NER and evaluated the quality of the predictions made by NeuroNER. Also, we have developed an algorithm that converts NeuroNER output to WebAnno input format. However, the performance metrics can be calculated and plotted by comparing the predicted labels with the gold labels. The evaluation can be done at the same time as the training if the test set is provided along with the training and validation sets, or separately after the training.