Source Critical Reinforcement Learning for Transferring Spoken Language Understanding to a New Language

Proceedings of the 27th International Conference on Computational Linguistics, pages 3597–3607 Santa Fe, New Mexico, USA, August 20-26, 2018. 3597 Source-Critical Reinforcement Learning for Transferring Spoken Language Understanding to a New Language He Bai1,2, Yu Zhou1,2, Jiajun Zhang1,2, Liang Zhao3, Mei-Yuh Hwang3 and Chengqing Zong1,2,4 1 National Laboratory of Pattern Recognition, Institute of Automation, CAS, Beijing, China 2 University of Chinese Academy of Sciences, Beijing, China 3 Mobvoi AI Lab, Seattle, US 4 CAS Center for Excellence in Brain Science and Intelligence Technology {he.bai, yzhou, jjzhang, cqzong}@nlpr.ia.ac.cn, {liangzhao, mhwang}@mobvoi.com Abstract To deploy a spoken language understanding (SLU) model to a new language, language transfer- ring is desired to avoid the trouble of acquiring and labeling a new big SLU corpus. Translating the original SLU corpus into the target language is an attractive strategy. However, SLU cor- pora consist of plenty of semantic labels (slots), which general-purpose translators cannot handle well, not to mention additional culture differences. This paper focuses on the language trans- ferring task given a tiny in-domain parallel SLU corpus. The in-domain parallel corpus can be used as the first adaptation on the general translator. But more importantly, we show how to use reinforcement learning (RL) to further finetune the adapted translator, where translated sen- tences with more proper slot tags receive higher rewards. We evaluate our approach on Chinese to English language transferring for SLU systems. The experimental results show that the gen- erated English SLU corpus via adaptation and reinforcement learning gives us over 97% in the slot F1 score and over 84% accuracy in domain classification. It demonstrates the effectiveness of the proposed language transferring method. Compared with naive translation, our proposed method improves domain classification accuracy by relatively 22%, and the slot filling F1 score by relatively more than 71%. 1 Introduction Spoken language understanding (SLU) is a key technique in today’s conversational systems such as Ap- ple Siri, Amazon Alexa and Microsoft Cortana. To make these conversational systems support multiple languages over different markets, collecting and annotating a large SLU training corpus per language are tedious and costly, and thus hinders the scalability of these systems. It would be greatly helpful if the efforts taken to develop one SLU system could be reused for other languages. For such a purpose, much work has been reported to explore language transferring of SLU systems or multilingual SLU systems (Garc´ıa et al., 2012; Calvo et al., 2013; Calvo et al., 2016; Jabaian et al., 2016). These systems can be grouped into two categories: test-on-source-model vs. train-on-target- language. Test-on-source-model is to translate the test sentence in the 2nd language (referred to as L2 from now on) into the first SLU system’s language (as L1), and then process it with the L1 SLU system. Train-on-target-language is to translate the L1 training corpus into L2, and then train an SLU system in L2. The train-on-target-language strategy allows tuning and adaptation of the models in the target language directly, and it avoids an overhead of machine translation during real-time execution. …

Content truncated to 500 words