A retrospective and future look at speech recognition application in assisting children with reading disabilities

• Main Authors: Husni, Husniza, Jamaludin, Zulikha
• Format: Book Section
• Language: English
• Published: WCWCS 2008 2008
• Subjects: QA76 Computer software
• Online Access: http://repo.uum.edu.my/1331/1/Husniza_husni_WCECS2008_pp555-558.pdf; http://repo.uum.edu.my/1331/; http://www.iaeng.org/publication/WCECS2008/WCECS2008_pp555-558.pdf
• Institution: Universiti Utara Malaysia

Parents of dyslexic children dream of computer technology and applications as panacea for their children. Dyslexic children’s deficits in phonological origin hinders reading skills to be developed at par with other ‘normal’ children. Thus this paper draws the various computer applications that have been used, and will probably be used in the future, to assist dyslexic children with reading disabilities.These applications often embed text-to-speech (TTS) technology and speech recognition (ASR) technology to aid children for both reading and writing. Given more attention to ASR, this paper reviews the previous success story of ASR, as well as future technology of it that will continue providing support for dyslexic children with diverse learning disabilities. In prospective, ASR is now advancing towards providing a reading tutor that could ‘listen’ to what is being read and provide correction to any incorrect reading. This paper highlights the potential of ASR in teaching the dyslexic children to read in Bahasa Melayu (Malaysia’s national language). For that purpose, ten dyslexic children were asked to read aloud 114 Bahasa Melayu words in seven different sessions and each speech was recorded, resulting in a total of 6834 utterances with 6323 utterances with errors. These utterances are transcribed and the pronunciations are modeled for ASR to train on.