Style and branding elements extraction from businessweb sites

Authors:
Limei Jiao;Suk Hwan Lim;Nina Bhatti;Yuhong Xiong;Jerry Liu
Affiliations:
Hewlett-Packard Laboratories, Beijing, China;Hewlett-Packard Laboratories, Palo Alto, CA, USA;Hewlett-Packard Laboratories, Palo Alto, CA, USA;Innovation Works, Beijing, China;Hewlett-Packard Laboratories, Palo Alto, CA, USA
Venue:
Proceedings of the 10th ACM symposium on Document engineering
Year:
2010

Citing 8
Cited 0

Identifying syntactic differences between two programs

Software—Practice & Experience
Web data extraction based on partial tree alignment

WWW '05 Proceedings of the 14th international conference on World Wide Web
Weighted Link Analysis for Logo and Trademark Image Retrieval on the Web

WI '05 Proceedings of the 2005 IEEE/WIC/ACM International Conference on Web Intelligence
Content-based multimedia information retrieval: State of the art and challenges

ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP)
Relevance feedback methods for logo and trademark image retrieval on the web

Proceedings of the 2006 ACM symposium on Applied computing
Joint optimization of wrapper generation and template detection

Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining
Searching for logo and trademark images on the web

Proceedings of the 6th ACM international conference on Image and video retrieval
LogoSeeker: a system for detecting and matching logos in natural images

Proceedings of the 15th international conference on Multimedia

Quantified Score

Hi-index	0.00

Visualization

Abstract

We describe a method to extract style and branding elements from multiple web pages in a given site for content repurposing. Style and branding elements convey the values of the site owners effectively and connect with the target prospects. They are manifested through logos, graphical elements, background color, font styles, font colors and other illustrations. Our method automatically extracts color and image elements appearing frequently and prominently on multiple pages throughout the site. We rely on a DOM tree matching method to obtain the frequency of re-occurring elements and use relative sizes and positions of elements to determine the type of elements. Note that approximate locations of these elements provide an added clue to the content repurposing engine as to where to place the elements in the repurposed document. The obtained results show that the proposed method can efficiently extract style and branding elements with high accuracy.