New data, benchmark and baseline for L2 speaking assessment for low-resoure languages
Kurimo, M., Getman, Y., Voskoboinik, E., Al-Ghezi, R., Kallio, H., Kuronen, M., von Zansen, A., Hilden, R., Kronholm, S., Huhta, A., & Linden, K. (2023). New data, benchmark and baseline for L2 speaking assessment for low-resoure languages. In Proceedings of the 9th Workshop on Speech and Language Technology in Education (SLaTE) (pp. 166-170). International Speech Communication Association. https://doi.org/10.21437/SLaTE.2023-32
DisciplineSoveltava kielentutkimusRuotsin kieliHyvinvoinnin tutkimuksen yhteisöSuomen kieliApplied language studiesSwedishSchool of WellbeingFinnish
© 2023 International Speech Communication Association
The development of large multilingual speech models provides the possibility to construct high-quality speech technology even for low-resource languages. In this paper, we present the speech data of L2 learners of Finnish and Finland Swedish that we have recently collected for training and evaluation of automatic speech recognition (ASR) and speaking assessment (ASA). It includes over 4000 recordings by over 300 students per language in short read-aloud and free-form tasks. The recordings have been manually transcribed and assessed for pronunciation, fluency, range, accuracy, task achievement, and a holistic proficiency level. We present also an ASR and ASA benchmarking setup we have constructed using this data and include results from our baseline systems built by fine-tuning self-supervised multilingual model for the target language. In addition to benchmarking, our baseline system can be used by L2 students and teachers for online self-training and evaluation of oral proficiency.
PublisherInternational Speech Communication Association
ConferenceWorkshop on Speech and Language Technology in Education
Is part of publicationProceedings of the 9th Workshop on Speech and Language Technology in Education (SLaTE)
Publication in research information system
MetadataShow full item record
Related funder(s)Academy of Finland
Funding program(s)Academy Project, AoF
Additional information about fundingThis work was done and the data were collected as part of the Academy of Finland grants number 322619, 322625, 322965 and 337073.
Showing items with similar title or keywords.
Tammelin-Laine, Taina (University of Jyväskylä, 2014)
Creaky voice and utterance fluency measures in predicting perceived fluency and oral proficiency of spontaneous L2 Finnish Kallio, Heini; Suviranta, Rosa; Kuronen, Mikko; Zansen, Anna von (International Speech Communication Association, 2022)While utterance fluency measures are often studied in rela tion to perceived L2 fluency and proficiency, the effect of creaky voice remains ignored. However, creaky voice is frequent in a number of languages, including ...
Inlärning och behärskning av svenskans verb- och adjektivböjning samt negationens placering hos finska grundskoleelever Paavilainen, Marika (University of Jyväskylä, 2015)
Prosody and fluency of Finland Swedish as a second language : Investigating global parameters for automated speaking assessment Kallio, Heini; Kautonen, Maria; Kuronen, Mikko (Elsevier BV, 2023)This study investigates prosody and fluency of Finland Swedish as a second language (L2). The main objective is to investigate global measures of prosody and fluency as predictors of overall oral proficiency, fluency, and ...
Kallio, Heini; Kuronen, Mikko (Guarant International, 2023)The aim of the study was to investigate whether integrating parameters based on pause location improve the prediction of fluency and proficiency in L2 Finnish monologic speech. To answer this question, multiple linear ...