1 / 23

Intelligent Interfaces Enabling Blind Users to Build Accessibility into the Web

Intelligent Interfaces Enabling Blind Users to Build Accessibility into the Web. A PhD Dissertation by Jeffrrey Bigham. Introduction. About the Author Remote Study of Blind Users’ Browsing Behavior AccessMonkey Audio Captcha Interface TrailBlazer WebAnywhere Conclusion.

misae
Download Presentation

Intelligent Interfaces Enabling Blind Users to Build Accessibility into the Web

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Intelligent Interfaces Enabling Blind Users to Build Accessibility into the Web A PhD Dissertation by JeffrreyBigham

  2. Introduction • About the Author • Remote Study of Blind Users’ Browsing Behavior • AccessMonkey • Audio Captcha Interface • TrailBlazer • WebAnywhere • Conclusion

  3. About the Author: Jeffrey Bigham • 2004-2009: CS PhD Student at UW • Advisor: Richard Ladner • 2009-2012: Assistant Professor at Rochester • Current: Carnegie Mellon • 35 under 35 MIT Innovator’s Award

  4. Rmoteweb study • Blind user group vs sighted user group • Comparative study with 10 ppl in each group • Remote due to difficulty of getting users into lab • Javascript plugin to monitor click events

  5. Results of remote web study • ~70% blind users to click on descriptive links vs 30% of sighted users • 20 times more dynamic content changes for sighted users • 20 times more interactions on average with dynamic content for sighted users • 3 times more probes per page on average for blind users • Only 6% blind users used skip links

  6. Criticisms • Study contained about half computer science and engineering people • But people with visual disabilities are underrepresented in computer science and engineering • Not as likely to find work arounds for flash • Only 6% blind users used skip links • Bigham hypothesis that screenreaers can jump around • More likely answer: skip links many times do not work, are very laggy, and can make the browser crash or freeze

  7. AccessMonkey • Transcodes webpages making them more accessible • FireFox plugin • Stand alone webpage • Web proxy

  8. How It Works • Adding alt text to images • Hand written by users • Grab text around image automatically • Intelligent links • Replace all graphics with links to the graphic • Add access keys to various links

  9. Results • IBM Japan and Google have created script repositories for AccessMonkey

  10. Criticisms • Only in javascript • Users must know how to write javascript • No studies done with blind users

  11. Audio Captcha: An Improved Interface • Audio captcha • Alternative to visual captcha for people who are blind • More difficult than visual captcha • Interfaces are terrible • Screenreaders talking while audio going at same time • Audio starts before able to type

  12. How It Works • Play/Pause/R/F inside edit box • Ready to type when audio starts playing • Use . For play/pause • Use / fast forwward • Use , rewind

  13. Results • 59% more likely for blind users to complete audio captcha on first try • No improvement in time to complete on successful attempt • Audio inherently takes longer than visual

  14. Criticisms • Not able to transcode new interface into all websites • Partly irrelevant as firefox plugin can bypass captcha • Interface not intuitive

  15. TrailBlazer • Knowledge base • How to perform common tasks on various websites • Similar KBs have been created for sighted users

  16. How it works • Directs focus to particular parts of the website • Each focused part is a step in the task, e.g. make airline reservations • Natural language interface: make reservations • If no KB entry for that site, then can try to use script from another similar site

  17. Results • Tasks performed on 15 most popular sites • 41% correct suggestion on first try • 70% correct suggestion in top 5 results • 5% correct on first try for human guessing

  18. Criticisms • No studies done with blind users • Small amount of how to scripts

  19. WebAnywhere • A screen reader built into a website • Can be accessed from any computer with sound and javascript • Does not need to be installed • Just go to website and start surfing the web with TTS

  20. How It Works • Reads left to right, top to bottom • Can skip between links, headers, tables, paragraphs, and other HTML elements • Review what was typed • Hit enter to follow links • Full functionality of a traditional screen reader

  21. Results • Most people found it tedious to use • But only slightly more tedious than a normal screen reader • Most said they would use the system in the future

  22. Criticisms • Difficult to change rate of speech • Internet delay limits speech rate • Only English version at time • Can only be used to read the web

  23. Conclusion • Web Accessibility • A difficult but important problem • Bigham offers four innovative solutions • WebAnywhere most successful and longest lasting • Audio captcha interface not used • AccessMonkey and TrailBlazer not well known but good apps

More Related