Multi-layer dialog generation for non-visual web access

Abstract

People with visual disabilities use screen-readers to browse the Web. The existing screen-readers have limited ways of presenting Web page content. I propose to turn non-visual Web browsing into a multi-layer mixed-initiative dialog-based interaction between users and computers. The suggested layers of dialog navigation are: basic screen-reading, DFS or BFS, and domain-specific. The support of adaptive dialogs is also planned. This research is aimed at improving and accelerating non-visual Web browsing for blind users.