Sequence labeling is a fundamental problem in machine learning, natural language processing and many other fields. A classic approach to sequence labeling is linear chain conditional random fields (CRFs). When combined with neural network encoders, they achieve very good performance in many sequence labeling tasks. One limitation of linear chain CRFs is their inability to model long-range dependencies between labels. High order CRFs extend linear chain CRFs by modeling dependencies no longer than their order, but the computational complexity grows exponentially in the order. In this paper, we propose the Neural Latent Dependency Model (NLDM) that models dependencies of arbitrary length between labels with a latent tree structure. We develop an end-to-end training algorithm and a polynomial-time inference algorithm of our model. We evaluate our model on both synthetic and real datasets and show that our model outperforms strong baselines.

翻译：序列标签是机器学习、自然语言处理和许多其他领域的一个基本问题。 序列标签的典型方法是线性链有条件随机字段(CRFs) 。 当与神经网络编码器相结合时,它们在许多序列标签任务中取得非常良好的性能。 线性链通用报告格式的一个局限性是它们无法模拟标签之间的长距离依赖性。 高顺序的通用报告格式通过模拟依赖性不长于其顺序,扩展线性链通用报告格式,但计算复杂性会按顺序急剧增长。 在本文中,我们提出了神经中端依赖性模型(NLDM),该模型可以模拟带有潜伏树结构的标签之间任意长度的依赖性。 我们开发了终端到终端培训算法和模型的多元时间推算法。 我们评估了我们的合成和真实数据集模型,并显示我们的模型超越了强大的基线。