Dear Samir, I think you can find some applications in the link to a paper below (that's what you're after, I suppose - in software the term 'use case' is usually defined in a special way). Your search might also include 'owl AND repositories' in databases as scholar or Springer.

BTW. Annotation is very useful for linked data on the Web. We don't want to fill a huge database to reason on, rather we use the description schema laid out in an ontology, for example.

http://ceur-ws.org/Vol-1014/paper_59.pdf

Dear Michael, Thanks for the link. Indeed, annotation is very useful for linked data on the Web. However, reasoning is the most important characteristic of semantic web language. Unfortunately, I did not fin a real use cas for a real life application (industrie for instance) where the reasoning is used. Prolog for instance, it is a logic programming language which widely used

I would indicate three basic types of use cases:

1) Querying data represented as graphs: http://en.wikipedia.org/wiki/Triplestore

2) Classification: assuming several classes are defined in an ontology, an individual with certain properties can be classified ( its membership of a certain class can be established)

3) Reasoning with rules (encoded in SWRL language). Typical reasoner, e.g. Pellet, is aware of taxonomic relations and can deteremine (i.e. assert) additional properties of an individual or relations between two individuals.

Many ideas and practical informations can be found in the book Semantic

Web Programming by Hebler, Fisher, et. al. : http://eu.wiley.com/WileyCDA/WileyTitle/productCd-047041801X.html

Certainly, OWL can be also used to define a schema, hence several use cases related to software development and integration can be found: schema matching, generation of a database structure, formalization of messages exchanged between agents, etc.

Just to make this thread searchable for the authors Piotr mentioned above, it's 'Hebeler', Fisher and others.

@Samir, if you're looking for industry ontologies, you might also be interested in applications of Gellish, which acts as a grammar and semantic modeling language for industrial products. They say: 'Development of the Gellish Modeling Method for knowledge representation, modeling of requirements, facilities, products and their properties and of Gellish and its English Dictionary, Taxonomy and Ontology and a universal database implementation method.'

You find them on Sourceforge, see below.

http://sourceforge.net/projects/gellish/

You can also take a look at the project ROCAS (Reasoning On the Cloud Applying Semantics)

https://code.google.com/p/rocas/

Dear all,

Most of the previous answers say that OWL seams the better option but I would like to open the discussion to other languages. For instance, the British Museum (London) is currently creating one of the most important databases with more than 2 million of resources of his museum and they use RDF as language. Following this link, http://tinyurl.com/ldhktge, the information about the tools used are explained.

Another example is ChEMBL database, which includes several databases about chemical and healthcare domains (https://www.ebi.ac.uk/chembl/downloads). The authors also use RDF to code more than 1 million of elements.

@David

This was the first use case I mentioned - querying triplestores. In fact, RDF constitute a set of triples (subject, predicate, object). At present triplestores can contain up to 300 bilion triples and can be efficiently queried with SPARQL language.

See also: http://www.w3.org/wiki/LargeTripleStores

OWL is built upon RDF. However, storing it in a database is not that efficient. Querying OWL model results in a bunch of small queries for triples. The best solution is to keep an ontology small and place the whole model into memory. It is a certain technological limit for industrial use cases for OWL.

Huh, that is quite open question, because there many hundreds of application scenarios around the Semantic Web. Maybe you can limit a little bit. The most practical applications where made with Linked-Open Data. Ontologies are more or less only used in the scientific environment, but they are mostly less practical (because of the very strict regulation).

You can use Semantic Web for generating explanations to statistical data.

http://www.ke.tu-darmstadt.de/resources/explain-a-lod

You’ll find applications for reason (more or less only possible with ontologies):

http://kaon2.semanticweb.org/

You’ll find semantic web applications which aims to visualize semantics data:

http://www.semavis.net

Or you can use the Semantic Web for Annotations:

http://www.corporate-semantic-web.de/html5-semantic-multimedia-presentation-tool.html