=Paper=
{{Paper
|id=None
|storemode=property
|title=Context in the Large: Contextual Issues when Dealing with Thousands of Ontologies
|pdfUrl=https://ceur-ws.org/Vol-626/invited.pdf
|volume=Vol-626
|dblpUrl=https://dblp.org/rec/conf/ekaw/Mottad10
}}
==Context in the Large: Contextual Issues when Dealing with Thousands of Ontologies==
<pdf width="1500px">https://ceur-ws.org/Vol-626/invited.pdf</pdf>
<pre>
Context in the Large: Contextual Issues when Dealing
           with Thousands of Ontologies


                      Enrico Motta1 and Mathieu d’Aquin1
      1
       Knowledge Media Institute, The Open University, Milton Keynes, UK
                e.motta@open.ac.uk, m.daquin@open.ac.uk




 Abstract: Back in 2005 we launched a research programme called “Next
 Generation Semantic Web Applications”, which pioneered the vision of
 exploiting the Semantic Web as a source of background knowledge to enable
 the development of a new class of intelligent applications. This research
 programme was based on some rather radical tenets, envisaging applications
 relying not simply on a few hand-picked ontologies, but on the dynamic
 selection and use of knowledge acquired from thousands of online ontologies .
 Five years later, this programme has produced a variety of applications, tools
 and techniques, concretely realizing the paradigm in scenarios such as ontology
 evolution, relation discovery, word sense disambiguation, and semantic
 enrichment of folksonomies.
     An important challenge when building applications that consume knowledge
 dynamically sourced from thousands of ontologies concerns dealing with the
 contextual nature of ontological resources. Specifically, new methods were
 needed to identify automatically whether the context of the original ontology
 still holds when the knowledge is reused in the application in hand.
     In addition, we have also looked at this issue in an application-independent
 way and carried out analytical studies that consider the Semantic Web itself as
 an object of study and attempt to characterize and understand the relations
 between the different epistemologies which are at the basis of the published
 conceptualizations. In particular, we have developed formal notions of
 agreement and disagreement between ontologies and used these to cluster
 ontologies which seem to share similar world views. To our knowledge these
 studies represent the first ever empirical analysis of large scale distributed
 conceptualizations and provide useful insights into the concrete practices used
 by ontology engineers. Specifically, these studies make it possible for us to see
 which contextual viewpoints tend to occur more frequently and which
 communities share specific conceptualizations.
     Finally, such results can be used further to develop highly structured
 ontology repositories, identifying and resolving contextual discrepancies
 between ontologies to facilitate and improve the efficiency of both automatic
 and manual access to resources on the Semantic Web.

</pre>