arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.08229cs.CLcs.AI
收藏

AIN：基于近似推理网络的快速精确序列标记

AIN: Fast and Accurate Sequence Labeling with Approximate Inference Network

School of Information Science and Technology, ShanghaiTech University（上海科技大学信息科学与技术学院）；Shanghai Engineering Research Center of Intelligent Vision and Imaging（上海市智能视觉与成像工程技术研究中心）；University of Chinese Academy of Sciences（中国科学院大学）；DAMO Academy, Alibaba Group（阿里巴巴达摩院）

Xinyu Wang, Yong Jiang, Nguyen Bach, Tao Wang, Zhongqiang Huang, Fei Huang, Kewei Tu

备注：
Accept to Main Conference of EMNLP 2020 (Short). Camera-ready, 8 Pages
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

摘要：

线性链条件随机场（CRF）模型是最广泛使用的神经序列标记方法之一。精确的概率推理算法，如前向-后向和维特比算法，通常应用于CRF模型的训练和预测阶段。然而，这些算法需要顺序计算，使得并行化是不可能的。在本文中，我们提出了采用并行近似变分推理算法的CRF模型。基于该算法，我们设计了一个近似推理网络，它可以与神经CRF模型的编码器连接，形成一个端到端的网络，该网络可以并行化，以实现更快的训练和预测。实验结果表明，我们提出的方法实现了12.7倍的改进，在解码速度与长句和竞争力的准确性相比，传统的CRF方法。

英文摘要：

The linear-chain Conditional Random Field (CRF) model is one of the most widely-used neural sequence labeling approaches. Exact probabilistic inference algorithms such as the forward-backward and Viterbi algorithms are typically applied in training and prediction stages of the CRF model. However, these algorithms require sequential computation that makes parallelization impossible. In this paper, we propose to employ a parallelizable approximate variational inference algorithm for the CRF model. Based on this algorithm, we design an approximate inference network that can be connected with the encoder of the neural CRF model to form an end-to-end network, which is amenable to parallelization for faster training and prediction. The empirical results show that our proposed approaches achieve a 12.7-fold improvement in decoding speed with long sentences and a competitive accuracy compared with the traditional CRF approach.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑