Coupling Distant Annotation and Adversarial Training for Cross-Domain Chinese Word Segmentation

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6662–6671 July 5 - 10, 2020. c⃝2020 Association for Computational Linguistics 6662 Coupling Distant Annotation and Adversarial Training for Cross-Domain Chinese Word Segmentation Ning Ding1,2 , Dingkun Long2, Guangwei Xu2, Muhua Zhu2, Pengjun Xie2, Xiaobin Wang2, Hai-Tao Zheng1∗ 1Tsinghua University, China 2Alibaba Group {dingn18}@mails.tsinghua.edu.cn, {zhumuhua}@gmail.com, {dingkun.ldk,kunka.xgw}@alibaba-inc.com, {chengchen.xpj,xuanjie.wxb}@alibaba-inc.com, {zheng.haitao}@sz.tsinghua.edu.cn Abstract Fully supervised neural approaches have achieved significant progress in the task of Chi- nese word segmentation (CWS). Nevertheless, the performance of supervised models tends to drop dramatically when they are applied to out- of-domain data. Performance degradation is caused by the distribution gap across domains and the out of vocabulary (OOV) problem. In order to simultaneously alleviate these two is- sues, this paper proposes to couple distant annotation and adversarial training for cross- domain CWS. For distant annotation, we re- think the essence of “Chinese words” and de- sign an automatic distant annotation mecha- nism that does not need any supervision or pre-defined dictionaries from the target do- main. The approach could effectively explore domain-specific words and distantly annotate the raw texts for the target domain. For ad- versarial training, we develop a sentence-level training procedure to perform noise reduction and maximum utilization of the source domain information. Experiments on multiple real- world datasets across various domains show the superiority and robustness of our model, significantly outperforming previous state-of- the-art cross-domain CWS methods. 1 Introduction Chinese is an ideographic language and lacks word delimiters between words in written sentences. Therefore, Chinese word segmentation (CWS) is often regarded as a prerequisite to downstream tasks in Chinese natural language processing. This task is conventionally formalized as a character- based sequence tagging problem (Peng et al., 2004), where each character is assigned a specific label to denote the position of the character in a word. With the development of deep learning techniques, recent years have also seen increasing interest in applying neural network models onto CWS (Cai ∗ Corresponding author Figure 1: Different word distributions for the newswire domain and the medical domain. and Zhao, 2016; Liu et al., 2016; Cai et al., 2017; Ma et al., 2018). These approaches have achieved significant progress on in-domain CWS tasks, but they still suffer from the cross-domain issue when they come to processing of out-of-domain data. Cross-domain CWS is exposed to two major challenges: 1) Gap of domain distributions. This is a common issue existing in all domain adapta- tion tasks. Source domain data and target domain data generally have different distributions. As a result, models built on source domain data tend to degrade performance when they are applied to tar- get domain data. Generally, we need some labeled target domain data to adapt source domain models, but it is expensive and time consuming to manu- ally craft such data. 2) Out of vocabulary (OOV) problem, which means there exist some words in the testing data that never occur in the training data. Source domain models have difficulties in recogniz- ing OOV words …

Content truncated to 500 words