Vision Based Page Segmentation Algorithm: Extended and Perceived Success

Abstract

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

A Quantitative Comparison of Semantic Web Page Segmentation Approaches

“…In (Akpinar and Yesilada 2012), the authors improve upon the popular VIPS algorithm. They focus on improving the first phase of VIPS, the visual block extraction.…”

Identifying Patterns in Eyetracking Scanpaths in Terms of Visual Elements of Web Pages

“…However, these existing approaches typically treat a web page as an image to identify these AoIs which means these scanpaths cannot be used to process web pages. In order to address this, our previous work automatically segments a web page and each segment becomes an AoI [5,6]. This allows relating AoIs with the underlying source code which is important for being able to process web pages by using the eyetracking data.…”

“…However, identifying visual elements and their role is the key for such reengineering process. To automatically process a web page and identify these elements, in our previous work we have extended and improved the Vision Based Page Segmentation (VIPS) algorithm [5,6]. This extended algorithm automatically discovers visual elements and relates them to the underlying source code.…”

“…This algorithm analyses and aggregates a group of people's eyetracking data to create a common scanpath in terms of visual elements of web pages (Section 3). Web pages are first automatically segmented into visual elements with the extended and improved version of the VIPS algorithm [6,5]. Eyetracking data is then exported and related to these visual elements.…”

Understanding Eye Tracking Data for Re-engineering Web Pages

“…String representations are created using the sequence of Areas of Interests (AoIs) that gets fixation [8]. Web page AoIs can be generated in different ways such as using grid-layout [8], the source code of Web pages [9] or the fixations' distribution over Web pages [10].…”