Iterative entity navigation via co-clustering semantic links and entity classes

Abstract

With the increasing volume of Linked Data, the diverse links and the large amount of linked entities make it difficult for users to traverse RDF data. As semantic links and classes of linked entities are two key aspects to help users navigate, clustering links and classes can offer effective ways of navigating over RDF data. In this paper, we propose a co-clustering approach to provide users with iterative entity navigation. It clusters both links and classes simultaneously utilizing both the relationship between link and class, and the intra-link relationship and intra-class relationship. We evaluate our approach on a real-world data set and the experimental results demonstrate the effectiveness of our approach. A user study is conducted on a prototype system to show that our approach provides useful support for iterative entity navigation.