Fine-grained Knowledge Fusion for Sequence Labeling Domain Adaptation
This work addresses domain adaptation for sequence labeling tasks, which is an incremental improvement over existing methods by considering multi-level domain relevance.
The paper tackles the problem of negative transfer in sequence labeling domain adaptation by addressing the diversity of individual target domain samples and elements within samples, proposing a fine-grained knowledge fusion model with domain relevance modeling. Experiments on three sequence labeling tasks show it outperforms strong baselines and state-of-the-art methods.
In sequence labeling, previous domain adaptation methods focus on the adaptation from the source domain to the entire target domain without considering the diversity of individual target domain samples, which may lead to negative transfer results for certain samples. Besides, an important characteristic of sequence labeling tasks is that different elements within a given sample may also have diverse domain relevance, which requires further consideration. To take the multi-level domain relevance discrepancy into account, in this paper, we propose a fine-grained knowledge fusion model with the domain relevance modeling scheme to control the balance between learning from the target domain data and learning from the source domain model. Experiments on three sequence labeling tasks show that our fine-grained knowledge fusion model outperforms strong baselines and other state-of-the-art sequence labeling domain adaptation methods.