MathSpeak: A non-ambiguous language for audio rendering of MathML

Abstract

For years, students with print disabilities have struggled to have access to instructional materials in the fields of science, technology, engineering and mathematics (STEM). People with print disabilities rely heavily on speech for information input. Mathematical expressions transmitted through typical spoken language are replete with multiple interpretations. These ambiguities create a substantial burden for the acquisition of basic mathematics and of knowledge from fields requiring a strong foundation in mathematics such as science, technology, and engineering. MathSpeak is based on a set of rules for conveying mathematical expressions in a non-ambiguous manner. The MathSpeak technology contains a computerised component that can easily and rapidly translate STEM materials into the non-ambiguous MathSpeak form, which can then be converted to an auditory rendering via a custom-designed high-quality computer-synthesised voice. This technology has great potential for increasing the accessibility to STEM materials and careers in related fields as shown by the efficacy studies.