Calibrating Structured Output Predictors for Natural Language Processing
This addresses the need for reliable confidence calibration in safety-critical NLP applications such as healthcare, though it is incremental as it builds on existing binary calibration methods.
The paper tackles the problem of calibrating confidence scores for output entities in NLP structured prediction models, proposing a general calibration scheme that outperforms current techniques and improves model performance across tasks like named entity recognition and question answering, with enhancements in out-of-domain scenarios.
We address the problem of calibrating prediction confidence for output entities of interest in natural language processing (NLP) applications. It is important that NLP applications such as named entity recognition and question answering produce calibrated confidence scores for their predictions, especially if the system is to be deployed in a safety-critical domain such as healthcare. However, the output space of such structured prediction models is often too large to adapt binary or multi-class calibration methods directly. In this study, we propose a general calibration scheme for output entities of interest in neural-network based structured prediction models. Our proposed method can be used with any binary class calibration scheme and a neural network model. Additionally, we show that our calibration method can also be used as an uncertainty-aware, entity-specific decoding step to improve the performance of the underlying model at no additional training cost or data requirements. We show that our method outperforms current calibration techniques for named-entity-recognition, part-of-speech and question answering. We also improve our model's performance from our decoding step across several tasks and benchmark datasets. Our method improves the calibration and model performance on out-of-domain test scenarios as well.