Bert Base Cased NER Reranker

🚀 bert-base-cased-NER-reranker

A BERT model designed for text ranking, trained on a synthetic literary NER context retrieval dataset to enhance NER performance.

🚀 Quick Start

To use this model, you need to construct a text in the form of NER-sentence [SEP] context-sentence. The model will predict the positive class if the context-sentence is useful for predicting the NER-sentence, and the negative class otherwise.

✨ Features

• Trained on Synthetic Dataset: This model is trained on the synthetic literary NER context retrieval dataset Amalvy et. al, 2023.
• High Performance: It achieves an F1 score of 98.34 on the synthetic test set.

📚 Documentation

Performance Metrics

The model attains an F1 score of 98.34 on the synthetic test set. For detailed information about NER performance gains when using this retriever model to assist a NER model at inference, refer to Amalvy et. al, 2023.

How to Reproduce Training

You can find the training script here.

📄 License

This project is licensed under the MIT license.

📖 Citation

If you use this model in your research, please cite the following:

@InProceedings{Amalvy2023,
  title	       = {Learning to Rank Context for Named Entity Recognition Using a Synthetic Dataset},
  author       = {Amalvy, A. and Labatut, V. and Dufour, R.},
  booktitle    = {2023 Conference on Empirical Methods in Natural Language Processing},
  year	       = {2023},
  doi	       = {10.18653/v1/2023.emnlp-main.642},
  pages	       = {10372-10382},
}