Semantic Web Dog Food

Presented at: 8th International Semantic Web Conference (ISWC2009)

by Christoph Boehm, Philip Groth, Ulf Leser

Ontologies are tools for describing and structuring knowledge, with many applications in searching and analyzing complex knowledge bases. Since building them manually is a costly process, there are various approaches for bootstrapping ontologies automatically through the analysis of appropriate documents. Such an analysis needs to find the concepts and the relationships that should form the ontology. However, since relationship extraction methods are imprecise and cannot homogeneously cover all concepts, the initial set of relationships is usually inconsistent and rather imbalanced - a problem which, to the best of our knowledge, was mostly ignored so far. In this paper, we define the problem of extracting a consistent as well as properly structured ontology from a set of inconsistent and heterogeneous relationships. Moreover, we propose and compare three graph-based methods for solving the ontology extraction problem. We extract relationships from a large-scale data set of more than 325K documents and evaluate our methods against a gold standard ontology comprising more than 12K relationships. Our study shows that an algorithm based on a modified formulation of the dominating set problem outperforms greedy methods.

Graph-Based Ontology Construction from Heterogenous Evidences was presented at this event.

Keywords: Semantic Web

Resource URI on the dog food server: http://data.semanticweb.org/conference/iswc/2009/paper/research/217

Explore this resource elsewhere:

• on Snorql
• on VisiNav
• on Sindice