论文标题
打破汉字的表示瓶颈：用中风序列建模的神经机器翻译
Breaking the Representation Bottleneck of Chinese Characters: Neural Machine Translation with Stroke Sequence Modeling
论文作者
论文摘要
现有研究通常将汉字视为代表性的最低单位。但是，这种汉字表现形式将遭受两个瓶颈：1）学习瓶颈，学习不能从其丰富的内部特征（例如激进分子和笔触）中受益； 2）参数瓶颈，每个字符必须由唯一的向量表示。在本文中，我们介绍了一种新颖的表现方法，用于汉字打破瓶颈，即Strokenet，它通过拉丁语中风序列（例如，“ ao1（凹入）”为“ ajaie”和“ ajaie”和“ ajaie”和“ tu1（convex）”）。具体而言，拼写物将每个中风都映射到特定的拉丁字符，从而允许类似的汉字具有相似的拉丁表示。通过引入Strokenet到神经机器翻译（NMT），现在可以完美地实现许多功能强大但不适用的技术（例如，共享子字词汇学习和基于密文的数据增强）。对广泛使用的NIST英语，WMT17中文英语和IWSLT17日语 - 英语NMT任务进行的实验表明，Strokenet可以为具有更少的模型参数提供明显的性能增强，而在强大的基础线上可以实现较少的模型参数，实现26.5 Bleu在WMT17中文任务上比以前不用任何报道的数据更好，而无需使用Monoling nol noroling数据。代码和脚本可在https://github.com/zjwang21/strokenet上免费获得。
Existing research generally treats Chinese character as a minimum unit for representation. However, such Chinese character representation will suffer two bottlenecks: 1) Learning bottleneck, the learning cannot benefit from its rich internal features (e.g., radicals and strokes); and 2) Parameter bottleneck, each individual character has to be represented by a unique vector. In this paper, we introduce a novel representation method for Chinese characters to break the bottlenecks, namely StrokeNet, which represents a Chinese character by a Latinized stroke sequence (e.g., "ao1 (concave)" to "ajaie" and "tu1 (convex)" to "aeaqe"). Specifically, StrokeNet maps each stroke to a specific Latin character, thus allowing similar Chinese characters to have similar Latin representations. With the introduction of StrokeNet to neural machine translation (NMT), many powerful but not applicable techniques to non-Latin languages (e.g., shared subword vocabulary learning and ciphertext-based data augmentation) can now be perfectly implemented. Experiments on the widely-used NIST Chinese-English, WMT17 Chinese-English and IWSLT17 Japanese-English NMT tasks show that StrokeNet can provide a significant performance boost over the strong baselines with fewer model parameters, achieving 26.5 BLEU on the WMT17 Chinese-English task which is better than any previously reported results without using monolingual data. Code and scripts are freely available at https://github.com/zjwang21/StrokeNet.
