Semantic Role Labeling as Dependency Parsing: Exploring Latent Tree Structures Inside Arguments
This work addresses a fundamental challenge in NLP by improving SRL accuracy, which is crucial for tasks like information extraction and machine translation, though it is incremental in nature.
The paper tackles the problem of semantic role labeling (SRL) by proposing to treat flat argument spans as latent subtrees, reducing SRL to a tree parsing task. The method achieves new state-of-the-art results on CoNLL05 and CoNLL12 benchmarks, outperforming all previous syntax-agnostic works.
Semantic role labeling (SRL) is a fundamental yet challenging task in the NLP community. Recent works of SRL mainly fall into two lines: 1) BIO-based; 2) span-based. Despite ubiquity, they share some intrinsic drawbacks of not considering internal argument structures, potentially hindering the model's expressiveness. The key challenge is arguments are flat structures, and there are no determined subtree realizations for words inside arguments. To remedy this, in this paper, we propose to regard flat argument spans as latent subtrees, accordingly reducing SRL to a tree parsing task. In particular, we equip our formulation with a novel span-constrained TreeCRF to make tree structures span-aware and further extend it to the second-order case. We conduct extensive experiments on CoNLL05 and CoNLL12 benchmarks. Results reveal that our methods perform favorably better than all previous syntax-agnostic works, achieving new state-of-the-art under both end-to-end and w/ gold predicates settings.