Web Page Transcoding Based on Eye Tracking

As web technologies evolve and enable designers to create more visually interactive and consequently complex web pages, unfortunately, accessing these pages in alternative forms such as by using screen read- ers or mobile devices, is still a major problem. In the context of eMINE project, an algorithm to visually identify visual elements and their roles in the web pages, and eMINE Scanpath Analysis Algorithm to generate a common scanpath for visual elements by using eye tracking data were proposed. In this technical report, a number of transcoding techniques and their architectures are discussed. These techniques aim to combine the visual element identification with eMINE Scanpath Analysis Algorithm.