One Model for All Domains: Collaborative Domain-Prefix Tuning for Cross-Domain NER

TL;DR

This paper proposes a novel collaborative domain-prefix tuning method for cross-domain NER that leverages frozen pre-trained language models and domain-specific prompts, improving transferability across multiple domains.

Contribution

It introduces a new text-to-text generative approach with domain-prefix tuning, enabling knowledge transfer without structural changes in PLMs for cross-domain NER.

Findings

Outperforms existing methods on Cross-NER benchmark

Effective in both single-source and multi-source transfer scenarios

Demonstrates flexible and robust domain adaptation

Abstract

Cross-domain NER is a challenging task to address the low-resource problem in practical scenarios. Previous typical solutions mainly obtain a NER model by pre-trained language models (PLMs) with data from a rich-resource domain and adapt it to the target domain. Owing to the mismatch issue among entity types in different domains, previous approaches normally tune all parameters of PLMs, ending up with an entirely new NER model for each domain. Moreover, current models only focus on leveraging knowledge in one general source domain while failing to successfully transfer knowledge from multiple sources to the target. To address these issues, we introduce Collaborative Domain-Prefix Tuning for cross-domain NER (CP-NER) based on text-to-text generative PLMs. Specifically, we present text-to-text generation grounding domain-related instructors to transfer knowledge to new domain NER tasks without structural modifications. We utilize frozen PLMs and conduct collaborative domain-prefix tuning to stimulate the potential of PLMs to handle NER tasks across various domains. Experimental results on the Cross-NER benchmark show that the proposed approach has flexible transfer ability and performs better on both one-source and multiple-source cross-domain NER tasks. Codes are available in https://github.com/zjunlp/DeepKE/tree/main/example/ner/cross.