提出可并行推断的序列标注模型，速度更快且约束条件下表现更优。
Bregman Conditional Random Fields: Sequence Labeling with Parallelizable Inference Algorithms
- 基于Bregman投影设计可并行推理的新模型
- 在强约束场景下优于均值场方法，速度更快
- 支持部分标签学习，适合高效序列标注任务
我们提出一种新的判别式序列标注模型——Bregman条件随机场（BCRF）。与标准线性链条件随机场不同，BCRF基于迭代Bregman投影，支持快速并行推理。我们展示了如何利用Fenchel-Young损失进行模型训练，包括从部分标签中学习的扩展。实验表明，该方法在性能上可媲美传统CRF，同时推理速度更快；在强约束条件下，优于另一种可并行化的均值场方法。
原文摘要 · Abstract (English)
We propose a novel discriminative model for sequence labeling called Bregman conditional random fields (BCRF). Contrary to standard linear-chain conditional random fields, BCRF allows fast parallelizable inference algorithms based on iterative Bregman projections. We show how such models can be learned using Fenchel-Young losses, including extension for learning from partial labels. Experimentally, our approach delivers comparable results to CRF while being faster, and achieves better results in highly constrained settings compared to mean field, another parallelizable alternative.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。