Content-based navigation within mathematical formulae on the web for blind users and its impact on expected user effort

Abstract

Learning Mathematics presents many challenges for blind students. The support for reading and navigating within webbased mathematical content is still limited in screen reader software, specially for languages such as Brazilian Portuguese. Lengthy and complex formulae, in particular, may demand significant effor from users who navigate only by keyboard and with linear feedback sound. Current screen readers provide limited support with semantic navigation based on content type. In this paper, we present a study involving the implementation of an initial prototype screen reader as proof-of-concept based on ChromeVox, aggregating new features for direct navigation with semantic mathematical elements, such as fractions, radicals and summation indexes. We also performed a study on the expected user effort of this approach, comparing it with the resources provided by other existing screen readers, namely JAWS, NVDA (Non-Visual Desktop Access). The estimation was performed using task models based on GOMS (Goals, Operators, Methods and Selection Rules) and KLM (Key-Level Model). The results showed that it is feasible to implement forms of navigation based on problem solving, because according to task models there was an expressive reduction in the estimated time to complete tasks.