JUCS - Journal of Universal Computer Science 17(7): 983-1004, doi: 10.3217/jucs-017-07-0983
Knowledge Extraction from RDF Data with Activation Patterns
expand article infoPeter Teufl‡, Günther Lackner§
‡ Graz University of Technology, Graz, Austria§ studio78.at, Graz, Austria

Corresponding author: Peter Teufl ( peter.teufl@iaik.tugraz.at )

© Peter Teufl, Günther Lackner.
This article is freely available under the J.UCS Open Content License.
Citation: Teufl P, Lackner G (2011) Knowledge Extraction from RDF Data with Activation Patterns. JUCS - Journal of Universal Computer Science 17(7): 983-1004. https://doi.org/10.3217/jucs-017-07-0983
Open Access
Abstract
RDF data can be analyzed with various query languages such as SPARQL. However, due to their nature these query languages do not support fuzzy queries that would allow us to extract a broad range of additional information. In this article we present a new method that transforms the information presented by subject-relationobject relations within RDF data into Activation Patterns. These patterns represent a common model that is the basis for a number of sophisticated analysis methods such as semantic relation analysis, semantic search queries, unsupervised clustering, supervised learning or anomaly detection. In this article, we explain the Activation Patterns concept and apply it to an RDF representation of the well known CIA World Factbook.
Keywords
machine learning, knowledge mining, semantic similarity, activation patterns, RDF, fuzzy queries