• Heidi Olivia Munksgaard
  • Dines Klewing Juul Madsen
This report details the development of tools to assist users in efficiently searching through large RDF datasets. To accomplish this, the research group has developed a graph visualisation application called GraphHelper, which can represent RDF-graphs from both local files and remote repositories. In addition the research group proposes two algorithm for intelligent searching in RDF-data.
Interesting Graph is an A*-like Breadth-First-Search algorithm, which to the best of the research groups knowledge, is a novel approach to exploring RDF-graphs. A user specifies search terms, based on which Interesting Graph searches the local neighbourhood of a designated node and returns the most promising sub-graph to be visualised by GraphHelper.
Interesting Path searches for relevant paths between two nodes, using a set of search terms specified by the user. This algorithm is an extension of previous work in the field. For both Interesting Graph and Paths, a set of experiments were performed in order to determine the precision of the algorithms using real-life data.
With this report comes a CD-ROM which contains documentation, source code, libraries and experimental results. GraphHelper and all libraries are open source.
SpecialiseringsretningIntelligente informationssystemer
SprogEngelsk
Udgivelsesdato1 aug. 2014
Antal sider142
ID: 201909283