PoolParty 2.7 offers new and comfortable ways to enrich any SKOS thesaurus with additional facts from the semantic web (see: LOD cloud). This functionality (which was extended significantly with version 2.7 in June 2010) supports any thesaurus manager to generate much richer knowledge models (ontologies) around specific domains than ever before (without facing high extra costs due to additional research). There are at least three arguments why one should consider building such “extended thesauri”:
- Use even more metadata to describe your resources and improve navigation and semantic search functionalities significantly
- Publish (at least) parts of your metadata / knowledge models as linked (open) data to stimulate innovative services around your contents on top of network effects
- Use linked data for data integration and semantic mashups; combine your own contents with contents from the web to improve your business intelligence
A short example: Just imagine you are working on a knowledge model in the area of “Skiing in Austria”. You have stated that one of Tyrols´s (most famous) skiing areas is “Kitzbühel“. By looking up at geonames.org you get extra metadata, e.g. alternate labels like “Kitzbichl” or longitude and latitude to display the concept on a map. In a next step you add famous Austrian skiers like “Hermann Maier” and “Franz Klammer“. From DBpedia you retrieve additional category information like Maier is a “Person born in 1972“, thumbnail pictures and also some links to other linked data sources, e.g. to the New York Times. Here we can learn that the NYT has mentioned Hermann Maier in 14 articles already. Finally we can add “Toni Sailer” as a third skier and we will find out by harvesting linked data that he was born in Tyrol and therefore we can add a new relation in our thesaurus between him and Tyrol.
We have learned: Linked Data can help us to build expressive knowledge models by using SKOS as an “interface” to the Linked Data Cloud.
SKOS thesauri can not only serve as a backbone for rich metadata structures to improve search applications but also as a new linked data source to be published and to be linked with other semantic data. PoolParty 2.7 follows many suggestions from “Linked Data Patterns” (edited by Leigh Dodds and Ian Davis) how linked data should be published. For instance, there are various ways with PoolParty 2.7 to identify resources, e.g. via “Patterned URIs” or via “Literal Keys” (see, for example, http://vocabulary.semantic-web.at/SemanticWebThesaurus/controlledvocabulary).
PoolParty uses TuQS as very fast linked data lookup service and can harvest data from virtually any linked (open) data source which provides a SPARQL-endpoint, e.g. DBpedia, Geonames, Wordnet, UMBEL or PoolParty sources themselves.
Don´t forget: SKOS stands for Simple Knowledge Organization System, thus PoolParty was designed as an easy-to-use Linked Data and Thesaurus Editor and Publishing System.