Ainsi, sans savoir ce que sera exactement le Web 3.0 on peut être certain qu'il sera sémantique, comme nous le font remarquer Technology Review ou le New York Times. Les technologies sont encore émergentes mais on peut imaginer que si des solutions simples permettaient de stocker, raisonner et retrouver des données sémantiques, il pourrait se démocratiser rapidement.

Bien des challenges restent néanmoins à traiter; mais avant même cela, certains se penchent sur le web 4.0. Il s'agit là d'un concept assez flou qui entend utiliser le web comme un immense réseau d'agents intelligents partageant leurs puissances de calculs sur cette base du web 3.0. Si l'objectif 3.0 s'approche, le 4.0 est encore loin, comme nous le montre la chronologie ci-dessous (cliquez pour agrandir l'image).

Soyez le premier à commenter cet article | Pages vues: 36336

With enthusiasm, J. Hendler described us what is going in this challenging area. What is particularly interesting is that we are really moving up in the semantic layer cake! Actually, since I started to work on this topic, I've observed that, year by year, the research focus is being smoothly raised in these layers.

So now the current stage is really higher than ontology, and even beyond the logic layer which is being intensively studied nowadays (notably with DL and Rules). So the hype wave is approaching the proof layer, which is an important layer for the acceptance of Semantic Web.

James gave an example based on transaction over the web: let's assume that a computer A send a bill to a computer B: "You owe me $20". Everybody expects the computer B won't give directly the money and say "Why? Prove it". And so the computer A could say: "you bought this book at this store and as you can see it is at this price". Additionally a rule saying that "when you buy an item you have to pay it" will finalize the proof.

The point is that all these pieces of information are given by RDF triples with references to some URI, so the computer B can check them and if it trusts the referenced site, he can accept the proof.

This example goes beyond some criticisms, notably coming the one from Google Executive, Peter Norvig, who challenged T. Berners-Lee during a conference. It is true that with less human oversight lot of people would attempt to exploit the Semantic Web to make SPAM of get better result to sell you pills of any colored and so it could be difficult to trust such a web. But such comment forgets the concept of proof. Actually what is important is not to say it is the truth, but to give the basis for the other side to know whether this can be the truth according to its beliefs, like we do intuitively when we manually collect data from various location. So have a good grounding is the key: we give the source of the information, so people can decide to trust it or not.

Some big challenges about managing security in such an open infrastructure are coming, but Semantic Web is getting closer.

Soyez le premier à commenter cet article | Pages vues: 18989

Je profite d’un article récent, The Semantic Web Revisited [2], pour présenter l’idée du SW, présentée initialement par Tim Berners-Lee en 2001 [1] dans le but d’enrichir l’information disponible sur internet pour pouvoir la rendre utilisable non seulement par les humains mais aussi par les ordinateurs.

Tout comme les physiciens du CERN avaient permis de porter le web vers le succès, l’article souligne que l’utilisation d’ontologies, un des concepts clés du Semantic Web, dans les e-sciences pourrait pousser l’adoption des technologies du web sémantique vers d’autres communautés (notamment les sciences environnementales) avant de s’étendre à toute la communauté internet.

Soyez le premier à commenter cet article | Pages vues: 25054