7th International Conference on Spoken Language Processing

September 16-20, 2002
Denver, Colorado, USA

Speech Recognition for Language Teaching and Evaluating: a Study of Existing Commercial Products

Rebecca Hincks

KTH, Sweden

Educators and researchers in the acquisition of L2 phonology have called for empirical assessment of the progress students make after using new methods for learning [1]. This study investigated whether unlimited access to a speech-recognition-based language learning program would improve the general goodness of pronunciation of a group of middle-aged immigrant professionals studying English in Sweden. Eleven students were given a copy of the program Talk to Me by Auralog as a supplement to a 200-hour course in Technical English, and were encouraged to practice on their home computers. Their development in spoken English was compared with a control group of fifteen students who did not receive software. Talk to Me uses speech recognition to provide conversational practice, visual feedback on prosody and scoring of pronunciation. A significant limitation of commercial systems currently available is their inability to diagnose specific articulatory problems. However, in this course students also met at regular intervals with a pronunciation tutor who could steer them in the right direction for finding the most important sections to practice for their particular problems. Students reported high satisfaction with the software and used it for an average of 12.5 hours. Students were pre- and post-tested with the automatic PhonePass SET-10 test from Ordinate Corp. Results indicate that practice with the program was beneficial to those students who began the course with a strong foreign accent but that students who began the course with intermediate pronunciation did not show the same improvement.


Full Paper

Bibliographic reference.  Hincks, Rebecca (2002): "Speech recognition for language teaching and evaluating: a study of existing commercial products", In ICSLP-2002, 733-736.