HLT4LL: Helmer Strik
Bio
Helmer Strik received his Ph.D. in Biophysics from the University of Nijmegen, where he is now associate professor in Speech Science and Technology. His research activities include speech production and perception, automatic speech recognition (ASR), spoken dialogue systems, and language and speech technology for computer assisted language learning (CALL), e-Learning, therapy, and e-Health. He has published over 200 refereed papers in international journals, books, magazines and proceedings, was invited as a panelist and keynote speaker at international conferences and symposia and has been guest editor for special issues of the Speech Communication journal. He is vice-chair of the scientific committee of the International Speech Communication Association Special Interest Group (ISCA SIG) on Speech and Language Technology in Education (SLaTE, www.sigslate.org), and a member of the international scientific committees of various speech-related conferences and workshops. At present, he coordinates several projects on e-Learning (CALL) and e-Health, some including serious gaming, in which dedicated Language and Speech Technology is being developed (see https://www.helmer-strik.nl/projects/). For instance, the European ‘Lifelong Learning Programme’ (LLP) e-Learning projects ‘Games Online for Basic Language learning’ (GOBL) and ‘Digital Literacy Instructor’ (DigLin), and the e-Health project ‘CHAllenging Speech training In Neurological patients by interactive Gaming’ (CHASING).
ASR-enabled CALL: challenges and opportunities
The growing interest in learning foreign languages across the world has led to an enormous increase in Computer Assisted Language Learning (CALL) programs that offer practice and automatic feedback for various language skills. In general, automatic systems that provide practice and feedback for oral skills are much less common. This has to do with the difficulties involved in automatically processing and analyzing the speech of second language learners. Applying Automatic Speech Recognition (ASR) to non-native speech is notoriously difficult and providing good-quality, immediate feedback on non-native spoken utterances is highly complex. This might be one of the reasons why ASR-enabled CALL does not have a long history, it started in the late 90s of the previous century.
Still, our research has shown that it is possible to develop CALL systems that employ ASR to provide practice and good-quality feedback on different aspects of second language speaking such as pronunciation, morphology and syntax. In addition, such CALL systems provide new opportunities for research, e.g. on second language learning (SLA). All interactions of the users with the system are stored, and by analyzing these log files behavioral data can be obtained. Such data are interesting for studying, e.g., the effects of factors such as motivation or different forms of practice and feedback on language learning, or to compare the effect of written vs. oral training. These and other topics have been studied in our research, and results will be presented.