Scalable Visualization of Ontologies using Power Law Graphs

Abstract

Semantic Web Visualization is a very important area of research because of its application in multiple domains such as analyzing topology of large and complex networks, semantic nets and more recently for ontologies. Ontology, as a formal explicit description of concepts in a domain of discourse, is one of the pillars in the Semantic Web. It is very important to visualize the semantic nets in order to understand their structure in a more understandable approach by the user. There is a growing need of e ective visualizations required during the design as well as maintenance phase of ontology. Several two-and three-dimensional visualization tools have been implemented and discussed in research literature. Di erent layouts and algorithmic approaches are followed in these tools. A tool may generate a drawing that makes it hard for user to understand the structure of a large semantic net. On the contrary, the representation may overlook the holistic view and thus losses the semantics in the ne grained associations between important concepts if the visualization focuses on the overview. A scalable solution is required to facilitate e ective visualization of very large semantic nets that could pave the way for a clearer understanding of the semantics and structure of the semantic nets. This thesis focus on developing a scalable solution to very large semantic net visualization problem and also to optimize the performance by applying ltering and clustering approaches on large scale graphs in order to get aesthetic visualization layout of semantic nets.