A Resource for Computational Experiments on Mapudungun

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 2872–2877 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 2872 A Resource for Computational Experiments on Mapudungun Mingjun Duan†1, Carlos Fasola†2, Sai Krishna Rallabandi1, Rodolfo M. Vega1, Antonios Anastasopoulos1, Lori Levin1, Alan W Black1 1Language Technologies Institute 2Interpreting and Translation Studies Carnegie Mellon University Wake Forest University 5000 Forbes Ave, Pittsburgh PA 15213, USA 1834 Wake Forest Road, Winston-Salem, NC 27109, USA {mingjund,srallaba,rmvega}@andrew.cmu.edu, fasolaca@wfu.edu, {aanastas, levin, awb}@cs.cmu.edu Abstract We present a resource for computational experiments on Mapudungun, a polysynthetic indigenous language spoken in Chile with upwards of 200 thousand speakers. We provide 142 hours of culturally significant conversations in the domain of medical treatment. The conversations are fully transcribed and translated into Spanish. The transcriptions also include annotations for code-switching and non-standard pronunciations. We also provide baseline results on three core NLP tasks: speech recognition, speech synthesis, and machine translation between Spanish and Mapudungun. We further explore other applications for which the corpus will be suitable, including the study of code-switching, historical orthography change, linguistic structure, and sociological and anthropological studies. Keywords: Mapudungun, Endangered Languages, Speech, Machine Translation 1. Introduction Recent years have seen unprecedented progress for Natural Language Processing (NLP) on almost every NLP subtask. Even though low-resource settings have also been explored, this progress has overwhelmingly been observed in lan- guages with significant data resources that can be leveraged to train deep neural networks. Low-resource languages still lag behind. Endangered languages pose an additional challenge. The process of documenting an endangered language typically includes the creation of word lists, audio and video record- ings, notes, or grammar fragments, with the created re- sources then stored in large online linguistics archives. This process is often hindered by the Transcription Bottleneck: the linguistic fieldworker and the language community may not have time to transcribe all of the recordings and may only transcribe segments that are linguistically salient for publication or culturally significant for the creation of com- munity resources. With this work we make publicly available a large corpus in Mapudungun, a language of the indigenous Mapuche people of southern Chile and western Argentina. We hope to ameliorate the resource gap and the transcription bottle- neck in two ways. First, we are providing a larger data set than has previously been available, and second, we are pro- viding baselines for NLP tasks (speech recognition, speech synthesis, and machine translation). In providing baselines and datasets splits, we hope to further facilitate research on low-resource NLP for this language through our data set. Research on low-resource speech recognition is par- ticularly important in relieving the transcription bottleneck, while tackling the research challenges that speech synthe- sis and machine translation pose for such languages could lead to such systems being deployed to serve more under- represented communities. †Equal contribution 2. The Mapudungun Language Mapudungun (iso 639-3: arn) is an indigenous language of the Americas spoken natively in Chile and Argentina, with an estimated 100 to …

Content truncated to 500 words