End-to-End Slot Alignment and Recognition for Cross-Lingual NLU

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 5052–5063, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 5052 End-to-End Slot Alignment and Recognition for Cross-Lingual NLU Weijia Xu∗ University of Maryland weijia@cs.umd.edu Batool Haider Amazon AI bhaider@amazon.com Saab Mansour Amazon AI saabm@amazon.com Abstract Natural language understanding (NLU) in the context of goal-oriented dialog systems typ- ically includes intent classification and slot labeling tasks. Existing methods to expand an NLU system to new languages use ma- chine translation with slot label projection from source to the translated utterances, and thus are sensitive to projection errors. In this work, we propose a novel end-to-end model that learns to align and predict target slot labels jointly for cross-lingual transfer. We introduce MultiATIS++, a new multilingual NLU corpus that extends the Multilingual ATIS corpus to nine languages across four language families, and evaluate our method using the corpus. Re- sults show that our method outperforms a sim- ple label projection method using fast-align on most languages, and achieves competitive per- formance to the more complex, state-of-the-art projection method with only half of the train- ing time. We release our MultiATIS++ corpus to the community to continue future research on cross-lingual NLU. 1 Introduction As a crucial component of goal oriented dialogue systems, natural language understanding (NLU) is responsible for parsing an utterance into a semantic frame to identify the user’s need. These semantic frames are structured by what the user intends to do (the intent) and the arguments of the intent (the slots) (Tur et al., 2010). Given the English example in Figure 1, we identify the intent of the utterance as “flight” and label the slots to extract the depar- ture city and airline name. Intent detection can be modeled as a sentence classification task where each utterance is labeled with an intent yI. Slot filling is typically modeled as a sequence labeling task where given the utterance x1...n, each word xi is labeled with a slot yi. ∗Work done at Amazon AI. Despite the high accuracy achieved by neural models on intent detection and slot filling (Goo et al., 2018; Qin et al., 2019), training such models on a new language requires additional efforts to collect large amounts of training data. One would consider transfer learning from high-resource to low-resource languages to minimize the efforts of data collection and annotation. However, currently available multilingual NLU datasets (Upadhyay et al., 2018; Schuster et al., 2019) only support three languages distributed in two language fami- lies, which hinders the study of cross-lingual trans- fer across a broad spectrum of language distances. In this paper, we release a new multilingual NLU corpus that contains training, development, and test data for six new languages in addition to the three languages in the Multilingual ATIS corpus (Upad- hyay et al., 2018). The resulting corpus, namely MultiATIS++, consists in total of 37,084 training examples and 7,859 test examples covering nine languages in four language families. Using our corpus, we explore the use of …

Content truncated to 500 words