[ Back to EurekAlert! ] Public release date: 16-Feb-2011
[ | E-mail Share Share ]

Contact: Shigeaki Sakurai
shigeaki.sakurai@toshiba.co.jp
Inderscience Publishers

Build your online networks using social annotations

Researchers at Toshiba are working on a way of finding clusters of like-minded bloggers and others online using "social annotations". Social annotations are the tags and keywords, the comments and feedback that users, both content creators and others associate with their content. Their three-step approach could help you home in on people in a particular area of expertise much more efficiently and reliably than simply following search engine results. The same tools might also be used in targeted marketing.

Users of photo gallery sites, such as Flickr (http://www.flickr.com/) and the social bookmarking sites like Delicious (http://delicious.com/) are familiar with social annotations. These services offer environments in which social annotations can be assigned with precision to different digital objects, whether a photograph, a bookmark, or more widely to blog posts, video uploads and discussion threads. With such annotations in place - notes and tags - users can easily search pictures or bookmarks. Toshiba's Shigeaki Sakurai and Hideki Tsutsui currently at Tokyo company NewsWatch suggest that the same social annotations could also be used to build networks.

It is relatively easy to create keyword-rich social annotations, the researchers explain, and such annotations represent an additional layer of knowledge concerning the digital objects with which they are associated. Previous researchers have considered how social annotations might be used to improve search engine results and how web pages might be automatically tagged based on such annotations.

The team explains that their technique allows them to divide bloggers into clusters according to their interests. The method calculates similarities between bloggers based on three steps: First it identifies target objects (subjects of interest), secondly it calculates similarities between target objects discussed in blog articles based on social annotations, and thirdly it calculates how these target objects are rated based on impression words in the blog posts. Products and services are considered the target objects in their study, which was tested on data from Commutents and the Yahoo Japan Movie services. These three steps are then combined to create cluster of bloggers.

###

"A clustering method of bloggers based on social annotations" in Int. J. Business Intelligence and Data Mining, 2011, 6, 26-41



[ Back to EurekAlert! ] [ | E-mail Share Share ]

 


AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert! system.