Vision Based Page Segmentation Algorithm: Extended and Perceived Success

摘要: Web pages consist of different visual segments, serving purposes. Typical structural segments are header, right or left columns and main content. Segments can also have nested structure which means some may include other segments. Understanding these is important in properly displaying web for small screen devices alternative forms such as audio reader users. There exist techniques identifying a page. One successful approach Vision Based Segmentation Algorithm (VIPS Algorithm) uses both the underlying source code rendering However, there limitations this paper explains how we extended improved VIPS built it Java. We conducted online user evaluations to investigate people perceive success segmentation granularity they prefer see page segmented. This presents preliminary results show that, with higher better regardless complexity.