Author and Advisors
Mater Thesis title: “Vocal Interaction for the navigation of mathematical expressions”
Author: Pier Felice Balestrucci.
Advisors: Alessandro Mazzei, Luca Anselma, Rossana Damiano.
External supervision by: doctor Cristian Bernareggi.
Assistive technologies are those technologies that make IT products, both hardware and software, accessible to people with impairment.
The goal of this Thesis is to make the listening of mathematical formulas easier and more intelligible for visually impaired people
Mathematical formulae are full of symbols that are hard to read for computer screen reader softwares.
Generally, whoever is affected by visual impairment can use a LaTeX representation of formulas in order to correctly read them via screen reader softwares;
that being said, LaTeX makes the reading verbose and slow, and results in being an obstacle for whoever does not know the LaTeX language.
The major goal of this Thesis is to produce a tool that can give visually impaired users advantages and simplifications.
This tool embodies two main functionalities:
- the translation of mathematical formulas into mathematical sentences, conducted via Natural Language Generation techniques
- the introduction of a dialog system that aims at navigating and exploring the given formula
This Thesis project is rooted in a previous work conducted by Michele Monticone, aiming at building a mathematical sentence generation system and the speech synthesis in the Italian language.
My work expands what was made by my colleague and is open to a wider audience, since it produces mathematical sentences in English; furthermore, it improves that work with a vocal assistant that helps the user to navigate the mathematical sentence and to understand it better, by
responding to specific needs such as the repetition of some parts of the sentence itself.