Skip to content

nicola883/web-segmentation-evaluation

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 

Repository files navigation

GroundTruth Web Page Segmentation Construction with MOB - Manual Design Of Blocks

In order to help human assessors building a Web page segmentation (ground truth) we developped the tool MoB (Manual-design-Of-Blocks). It is designed as a browser extension and offers functionalities to expert users for creating a manual segmentations. Users can create blocks based on Web page elements. They can merge blocks, navigate into the element hierarchy to produce a block graph, or produce a flat segmentation (i.e. leaves in the block tree). These segmentations are stored in arepository for the evaluation.

Web Page Segmentation with Block-o-Matic (BoM)

Our segmenter BoM segments a Web page without having a priori knowledge of its content and using only the heuristic rules defined by the W3C Web standards. For instance, we detect blocks using HTML5 content categories instead of using the tag names or text features. The segmentation process is composed of two main phases : detecting fine-grained (as small as possible) blocks and then merging them according to a stop condition, so that the segmentation is performed at the desired granularity. It can be expressed by the following function Φ(W), where W is the rendered DOM of a Web page.

More info about BoM

Links

Download MOB from Here

For installing the extension follow the same process as the BOM extension installation instructions

Download the Quick guide

Dataset: Web pages sources (in ScrapBook firefox extension format)

About

Tools for web page segmentation evaluation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 70.4%
  • CSS 25.4%
  • JavaScript 3.8%
  • ColdFusion 0.4%