Exploring Web Resources in Social Annotation Services in Hierarchy

This paper considers the problem of the organizing way of Web resources in social annotation systems. Currently, the Web resources in social annotation systems contain multiple topics, organizing them into different semantic clusters will facilitate userspsila browsing and search greatly. This paper proposes a hierarchy clustering method to organize Web resources in hierarchy by making full use of the linkage relationship between the Web resources and their annotations. In this way, users can browse or search a Web resource collection in a Google mappsilas way by going deep into a collection with more details. Some experiments on data set from Deli.icio.us are performed to justify the effectiveness of our method and a prototype based on pictures in Flickr is developed to demonstrate our idea.