arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1902.10985cs.CL
收藏

更好、更快、更强的序列标记组成解析器

Better, Faster, Stronger Sequence Tagging Constituent Parsers

Universidade da Coruña, CITIC(拉科鲁尼亚大学,CITIC);University of Copenhagen(哥本哈根大学)

David Vilares, Mostafa Abdou, Anders Søgaard

摘要:

用于成分分析的序列标记模型比其他类型的分析器更快,但准确性较差。在这项工作中,我们解决了以下弱点,这样的成分解析器:(a)高出错率关闭括号长成分,(b)大标签集,导致稀疏性,和(c)错误传播所产生的贪心解码。为了有效地关闭括号,我们训练了一个学习在标记方案之间切换的模型。为了减少稀疏性,我们分解标签集,并使用多任务学习来联合学习预测子标签。最后,我们通过辅助损失和带有策略梯度的缓存级微调来缓解贪心解码的问题。结合这些技术,我们显然超过了英语和中文Penn Treebanks上的序列标记成分解析器的性能,并进一步减少了它们的解析时间。在SPMRL数据集上,我们观察到全面的更大改进,包括巴斯克语,希伯来语,波兰语和瑞典语的最新技术水平。

英文摘要:

Sequence tagging models for constituent parsing are faster, but less accurate than other types of parsers. In this work, we address the following weaknesses of such constituent parsers: (a) high error rates around closing brackets of long constituents, (b) large label sets, leading to sparsity, and (c) error propagation arising from greedy decoding. To effectively close brackets, we train a model that learns to switch between tagging schemes. To reduce sparsity, we decompose the label set and use multi-task learning to jointly learn to predict sublabels. Finally, we mitigate issues from greedy decoding through auxiliary losses and sentence-level fine-tuning with policy gradient. Combining these techniques, we clearly surpass the performance of sequence tagging constituent parsers on the English and Chinese Penn Treebanks, and reduce their parsing time even further. On the SPMRL datasets, we observe even greater improvements across the board, including a new state of the art on Basque, Hebrew, Polish and Swedish.

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。

↑