EAI Endorsed Transactions on Ambient Systems 17(13): e3

Research Article

Enabling Text-To-Speech Functionality for Websites and Applications Using a Content-Derived Model

Download34 downloads
  • @ARTICLE{10.4108/eai.17-5-2017.152547,
        author={R. Soic and M. Vukovic and Z. Car},
        title={Enabling Text-To-Speech Functionality for Websites and Applications Using a Content-Derived Model},
        journal={EAI Endorsed Transactions on Ambient Systems},
        volume={17},
        number={13},
        publisher={EAI},
        journal_a={AMSYS},
        year={2017},
        month={5},
        keywords={text-to-speech, website content analysis, website structure, web content presentation, web accessibility, visual impairments},
        doi={10.4108/eai.17-5-2017.152547}
    }
    
  • R. Soic
    M. Vukovic
    Z. Car
    Year: 2017
    Enabling Text-To-Speech Functionality for Websites and Applications Using a Content-Derived Model
    AMSYS
    EAI
    DOI: 10.4108/eai.17-5-2017.152547
R. Soic1, M. Vukovic1,*, Z. Car1
  • 1: University of Zagreb Faculty of Electrical Engineering and Computing, Unska 3, Zagreb, Croatia
*Contact email: marin.vukovic@fer.hr

Abstract

Today’s society is largely dependent on the Internet websites that provide it with access to various information. While most users take accessing information in this way as granted, some user groups have significant issues while trying to access websites by traditional means. In this context, we focus on persons with severe visual impairment that are unable to use websites that are not text-to-speech enabled. The paper analyses existing solutions and proposes a model for enabling text-to-speech functionality based on deriving tree-like structure from website or web application content. The proposed model is evaluated on two case-studies; institutional website and interactive educational web application.