Mining Periodic Event Patterns from RDF Datasets

  • Anh Le
  • Michael Gertz
Conference paper

DOI: 10.1007/978-3-642-40683-6_13

Volume 8133 of the book series Lecture Notes in Computer Science (LNCS)
Cite this paper as:
Le A., Gertz M. (2013) Mining Periodic Event Patterns from RDF Datasets. In: Catania B., Guerrini G., Pokorný J. (eds) Advances in Databases and Information Systems. ADBIS 2013. Lecture Notes in Computer Science, vol 8133. Springer, Berlin, Heidelberg

Abstract

Exposing and sharing data and information using linked data sources is becoming a major theme on the Web. Several approaches have been developed to model and efficiently query and match linked open data, primarily represented as RDF graphs from RDF facts and associated ontological frameworks. Interestingly, little work has yet been conducted to discover interesting patterns from such data.

In this paper, we present an approach that aims at discovering interesting periodic event patterns from RDF facts describing events, for example, music events or festivals. Our focus is on exploiting the temporal and geographic properties associated with such event descriptions as well as the concept hierarchies used to categorize the different components of event facts. Discovered patterns of periodic events can be used for prediction or detection of outliers in RDF datasets. We demonstrate the feasibility and utility of our framework using real event datasets extracted as RDF facts from the Website eventful.com.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Anh Le
    • 1
  • Michael Gertz
    • 1
  1. 1.Database Systems Research GroupHeidelberg UniversityGermany