The label annotation mistakes by human annotators brings up two challenges to NER:

• mistakes in the test set can interfere the evaluation results and even lead to an inaccurate assessment of model performance.

• mistakes in the training set can hurt NER model training.

  Addressing these two problems :

• manually correcting the mistakes in the test set to form a cleaner benchmark.

• develop framework CrossWeigh to handle the mistakes in the training set.

CrossWeigh

Content

CrossWeigh works with any NER algorithm that accepts weighted training instances. It is composed of two modules. 1) mistake estimation: where potential mistakes are identified in the training data through a cross-checking process and 2) mistake re-weighing: where weights of those mistakes are lowered during training the final NER model.

Acknowledgements

Named-Entity-Recognition-NER-Papers
Pengfei Liu, Jinlan Fu and other contributors.

Inspiration

Name Entity Recognition