VerifiNER: Verification-augmented NER via Knowledge-grounded Reasoning with Large Language Models
This addresses the issue of lack of faithfulness in NER predictions for domain-specific applications like biomedical NER, though it is incremental as it builds on existing methods with a verification step.
The paper tackles the problem of erroneous predictions in domain-specific named entity recognition (NER) by proposing VerifiNER, a post-hoc verification framework that uses knowledge and large language models to identify and revise errors, resulting in improved faithfulness in predictions on biomedical datasets.
Recent approaches in domain-specific named entity recognition (NER), such as biomedical NER, have shown remarkable advances. However, they still lack of faithfulness, producing erroneous predictions. We assume that knowledge of entities can be useful in verifying the correctness of the predictions. Despite the usefulness of knowledge, resolving such errors with knowledge is nontrivial, since the knowledge itself does not directly indicate the ground-truth label. To this end, we propose VerifiNER, a post-hoc verification framework that identifies errors from existing NER methods using knowledge and revises them into more faithful predictions. Our framework leverages the reasoning abilities of large language models to adequately ground on knowledge and the contextual information in the verification process. We validate effectiveness of VerifiNER through extensive experiments on biomedical datasets. The results suggest that VerifiNER can successfully verify errors from existing models as a model-agnostic approach. Further analyses on out-of-domain and low-resource settings show the usefulness of VerifiNER on real-world applications.