We present Paged Graph Visualization (PGV), a new semi-autonomous tool for RDF data exploration and visualization. PGV consists of two main components: a) the “PGV explorer” and b) the “RDF pager” module utilizing BRAHMS, our high per-formance main-memory RDF storage system. Unlike existing graph visualization techniques which attempt to display the entire graph and then filter out irrelevant data, PGV begins with a small graph and provides the tools to incrementally explore and visual-ize relevant data of very large RDF ontologies. We implemented several techniques to visualize and explore hot spots in the graph, i.e. nodes with large numbers of immediate neighbors. In re-sponse to the user-controlled, semantics-driven direction of the exploration, the PGV explorer obtains the necessary sub-graphs from the RDF pager and enables their incremental visualization leaving the previously laid out sub-graphs intact. We outline the problem of visualizing large RDF data sets, discuss our interface and its implementation, and through a controlled experiment we show the benefits of PGV.

Publication

Leonidas Deligiannidis, Krys J. Kochut, and Amit P. Sheth, "RDF Data Exploration and Visualization". ACM CIKM 2007, First Workshop on CyberInfrastructure: Information Management in eScience (CIMS). Lisboa, Portugal Nov. 9th 2007. pp39-46. link to paper in pdf

Specifications

Installation

  1. Java 1.5
    • Get the JDK and install it. Reminder: set up the PATH environment variable.
  2. GraphViz
    • Simply download and execute, path gets set up.
  3. Sphinx4 Speech Recognizer
    • Download Sphinx-4 and unzip it.
    • The jar files that you need are: ONE and TWO.
  4. FreeTTS 1.2 Speech Synthesizer
    • Download and unzip the bin distribution.
    • Go into the lib directory and execute jsapi.exe and accept the licence.
    • Copy the speech.properties file to your home directory.
    • The jar files that you need are: lib/freetts.jar and lib/jsapi.jar.
  5. PGV source code comming soon
    • Modify the PGV.properties file.
    • j to compile and jrun to execute.
    • Javadoc draft