Cross-Lingual Dependency Parsing with Late Decoding for Truly Low-Resource Languages
Cross-Lingual Dependency Parsing with Late Decoding for Truly Low-Resource Languages
复制标题
DOI：
10.18653/v1/e17-1021
复制
发表时间：
2017-01
期刊：
影响因子：
--
通讯作者：
M. Schlichtkrull;Anders Søgaard
中科院分区：
文献类型：
--
作者：
M. Schlichtkrull;Anders Søgaard
在跨语言依存注释投影中，由于早期解码，信息在传输过程中经常丢失。我们提出了一种基于端到端图的神经网络依赖解析器，可以训练它来重现边缘分数矩阵，该矩阵可以直接投影到单词对齐上。我们表明，与之前的最佳技术水平相比，我们的跨语言依存解析方法不仅更简单，而且在 10 种语言中平均实现了 2.25% 的绝对改进。
In cross-lingual dependency annotation projection, information is often lost during transfer because of early decoding. We present an end-to-end graph-based neural network dependency parser that can be trained to reproduce matrices of edge scores, which can be directly projected across word alignments. We show that our approach to cross-lingual dependency parsing is not only simpler, but also achieves an absolute improvement of 2.25% averaged across 10 languages compared to the previous state of the art.