The use of headings in web pages can make pages more accessible for blind users. Unfortunately, many web pages do not use proper HTML markup (h1-h6 tags) to indicate headings, instead using visual styling stylings to create headings, thus making the distinction between headings and other page text invisible to blind users. We have created a classifier, HeadingHunter, that predicts whether web page text semantically functions as a heading by examining visual features of the text as rendered in a web browser. We conducted a user study where sighted people labeled what they perceived to be headings on a set of web pages. While we primarily performed this study to generate good test data for evaluating the performance of HeadingHunter, we also observed interesting labeling behavior from the participants. We are also conducting a second user study to observe the performance of blind participants on different versions of popular web pages with the HTML headings (1) removed, (2) as originally labeled, (3) as predicted by HeadingHunter, and (4) as labeled by a majority of participants from the first user study.
Advised by Richard Ladner
CSE 203
Wednesday
May 7, 2008
3:30 - 4:20 pm