Document Type

Project

Lead Author Type

CIS Masters Student

Advisors

Dr. Yonglei Tao, taoy@gvsu.edu

Embargo Period

3-12-2014

Abstract

Speech applications are currently used successfully in many contexts in which the auditory modality allows users to accomplish tasks efficiently and effectively, such as the automation of banking services normally handled by live agents. Such technology is also increasingly being utilized for language education to teach various aspects of the target language. However, though games are routinely played in language classrooms to help students practice elements of spoken language such as pronunciation and conversational skills, educational speech game software is still not widely available. This project explores technical issues involved in adapting some of these classroom games for inclusion in a voice-controlled application for students of English as a Second Language. Games on spelling and sentence formation in English were designed and implemented as part of a multimodal prototype, Let’s Speak!. This system, built using Microsoft’s speech recognition namespaces, combines the visual and auditory modalities in order to overcome problems that each modality on its own typically poses to users. The results seen here demonstrate that developing simple, yet engaging speech games for students is certainly possible using existing technologies. Furthermore, specific design and implementation issues encountered here can shed light on effective Voice User Interface design for future efforts in educational speech application development.

Share

COinS