+> Modern websites are becoming larger and more complex, making complete accessibility evaluations impractical due to the vast number of pages and page states. Traditional assessment methods often rely on ad-hoc or manual sampling, which can result in evaluations that lack representativeness and fail to effectively identify accessibility barriers. This work introduces a data-driven methodology that enhances web accessibility evaluation by systematically selecting representative pages. By applying clustering and unsupervised learning to group similar pages, the approach enables structured sampling and improves the accuracy and efficiency of accessibility audits.
0 commit comments