Multilingual and Cross-Lingual Intent Detection from Spoken Data

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7468–7475 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 7468 Multilingual and Cross-Lingual Intent Detection from Spoken Data Daniela Gerz∗, Pei-Hao Su,∗ Razvan Kusztos, Avishek Mondal, Michał Lis, Eshan Singhal, Nikola Mrkši´c, Tsung-Hsien Wen, and Ivan Vuli´c PolyAI Limited London, United Kingdom {dan,eddy,ivan}@polyai.com Abstract We present a systematic study on multilingual and cross-lingual intent detection (ID) from spoken data. The study leverages a new re- source put forth in this work, termed MINDS- 14, a first training and evaluation resource for the ID task with spoken data. It covers 14 in- tents extracted from a commercial system in the e-banking domain, associated with spoken examples in 14 diverse language varieties. Our key results indicate that combining machine translation models with state-of-the-art multi- lingual sentence encoders (e.g., LaBSE) yield strong intent detectors in the majority of tar- get languages covered in MINDS-14, and of- fer comparative analyses across different axes: e.g., translation direction, impact of speech recognition, data augmentation from a related domain. We see this work as an important step towards more inclusive development and eval- uation of multilingual ID from spoken data, hopefully in a much wider spectrum of lan- guages compared to prior work. 1 Introduction and Motivation A crucial functionality of Natural Language Un- derstanding (NLU) components in task-oriented dialogue systems is intent detection (ID) (Young et al., 2002; Tür et al., 2010; Coucke et al., 2018). In order to understand the user’s current goal, the system must classify their utterance into several predefined classes termed intents.1 Scaling dialogue systems in general and intent detectors in particular to support a multitude of new dialogue tasks and domains is a challenging, time- consuming, and resource-intensive process (Wen et al., 2017; Rastogi et al., 2019). This problem is further exacerbated in multilingual setups: it is ∗Both authors equally contributed to this work. 1For instance, in the banking domain utterances referring to cash withdrawal or currency exchange rates should be classified to the respective intent classes (Casanueva et al., 2020). An error in intent detection is typically the first point of failure for any task-oriented dialogue system. extremely expensive to annotate sufficient task data in each of more than 7,000 languages (Bellomaria et al., 2019; Xu et al., 2020). As a consequence, the current ID work has been largely constrained only to English, and standard ID benchmarks also exist only in English (Hemphill et al., 1990; Lar- son et al., 2019; Liu et al., 2019b; Casanueva et al., 2020; Larson et al., 2020, inter alia). The need to widen the reach of dialogue technology to other languages has been recognised only recently, and thus even text-based multilingual ID datasets are still few and far between: Schuster et al. (2019) provide NLU data in three languages (English, Spanish, Thai), while a more recent MultiATIS++ dataset (Xu et al., 2020) manually translates the ATIS dataset (Hemphill et al., 1990) from English to 8 target languages, extending the work of …

Content truncated to 500 words