The semantic web can be described as a web for machines. For that purpose, it requires expression of formalised knowledge on the web (in languages like RDF). This is aimed at bringing more precision to the knowledge gathered from the web than automatically indexing text documents. However, the semantic web suffers from a bootstrap problem: it can only start providing benefits when there is enough knowledge available and people will not provide knowledge if this does not return benefits.
In order to overcome this problem, we want to provide tools that allow people to start using semantic web technologies locally, for personal purpose and personal benefits, and that can spread through global and social communication.
Our working example is the use of semantic web technologies for annotating resources (and particularly pictures) on one's computer. This is already a very helpful application since it can help people querying their data on very fine and personal vocabulary (this is more powerful than simple schemes such as those used in flikr). We have developed a picture annotation editor that allow people to extend ontologies on the fly as they annotate them with very limited overhead.
The next step in such applications is to share these resources in a socially organised way, i.e., to friends, family, colleagues or the entire world. For that purpose, we just make use of the kind of architechture that has already been successful for sharing: peer-to-peer networks.
However, because the annotations have been made in some personnal ontology, which is not shared by everyone, it is necessary to be able to match the ontologies in order to share resources among peers (either by exchanging annotations or by querying other peers). We have designed services for matching ontologies and using the obtained alignment querying resources annotated by others.
We want to progress further in that direction by helping peers to interact through the ontology alignments. In particular, we would like to investigate the following directions:
The goal of this doctoral work is to investigate the above mentionned issues and to propose some designs for supporting these features. The two main topics will be the reinforcement of knowledge and network analysis for promoting the collaboration between peers. This requires addressing the problem of analysing adequately the network of peers while maintaining and acceptable degree of privacy.
Our prototype applications, as well as available tools for ontology matching will have to be enhanced in order to experimentally validate the design issues (we already have some experimental data that can be used).
Doctoral school: École doctorale MSTII, Grenoble.
Advisor: Jérôme Euzenat (Jerome:Euzenat#inrialpes:fr)
Group: Équipe Exmo, INRIA Rhône-Alpes
Other offers and administrative information