Institute for Informatics
Georg-August-Universität Göttingen

Databases and Information Systems

dbis
Uni Göttingen

2nd Querying Graph Structured Data (GraphQ 2010),
Novi Sad, Serbia, September 20, 2010. CEUR Workshop Proceedings 639, pp. 1-10.

A Configurable Graph Data Type for Online Exploration of Web Data

Thomas Hornung, Wolfgang May, Daniel Schubert

Abstract:

Application domains often include notions that are inherently based on graph structures. In this paper, we propose a comprehensive generic ontology-based datatype for graphs. It focuses on those aspects of graphs that are useful for workflows that require exploration of relevant parts of potentially large graphs by online algorithms. The goal of the ontology is to include as much information as possible supporting the graph exploration process declaratively into the specification of the graph. This allows to separate the (also declarative) specification of the actual exploration process from the maintenance of the graph itself. For concrete applications, the graph specification is given in RDF using this ontology. From the specification, an appropriate instantiation of the abstract datatype is automatically derived which is then used in information workflows.