ACup: Accessibility Catch-Up

Can a framework be developed so that accessibility research can be passed on to users more quickly?

Project summary 

Many people who access the Web need to do so in a way that doesn’t fit the standard model of someone sitting in a quiet office using a desktop computer – they may be in a noisy or visually difficult environment, or using a small screen on a mobile device, or they might have a visual impairment that necessitates the use of a screen reader, which uses synthetic speech to vocalise the page.  There are techniques which can be used (and should be used, but often aren’t)  when writing Web pages to make them more accessible to these users.  The ACup project explores whether it is possible to build a framework so that Web pages can be modified in the browser to add accessibility features, either standard ones that have been omitted, or novel accessibility functionality that has only just emerged from research.  This will enable more people to be able to access more Web content in a more effective and efficient way.

Further Information

Full scientific details:
Code repository:
Data repository:
Technical reports:
Funded by: This research was funded by a Google Research Award.
This project is in progress

Expanded Details

Our SASWAT project built a model of how sighted users interact with dynamic Web Content (‘Web 2.0′), and used this model to design a set of rules for presenting updates to screen-reader users. These were tested using a prototypical screen-reader, and found to be successful. This approach, however, while helpful for academic proof-of-concept testing, does not enable the visually impaired community to benefit quickly from results of the research, and led to us wondering if our findings, and the results of similar research, could be passed on more quickly to the end-users.

The Accessibility Catch-Up (ACup) project is a project, funded by a Google Research Award, investigating this problem. It builds upon the Google AxsJAX framework, which allows accessibility features to be automatically injected into Web pages. The AxsJAX framework, however, requires knowledge about the structure of a specific page; we were interested to find out if a more generic framework can be developed to inject code into any Web page that will allow any updates to be classified, then presented and interacted with in an appropriate manner, based on the SASWAT findings, for users with any ARIA-aware screen reader?

Completion Information

Final Report Summary:

This project is still in progress.

Completion date: 
Final report: Pending

▲ Up to the top