Re: Is LOD-cloud alive?

Hi Krzysztof,
Thanks for the clarification. Please find below some few comments...
> 
> 
> Sorry for not being clear enough. To the best of my understanding (and by trying myself) you cannot simply submit your ontology to LOV and it will be listed there. The LOV team will, for instance, ask you to align your ontology to other ontologies.

No, the team will ask you to have minimum metadata and labels, with link to existing vocabulary. That’s why you have the “L” in LOV :)
> This may turn out to be difficult for various reasons. For example,  for legal reasons companies that develop ontologies sometimes do not want to have relations to other ontologies in their own ontology especially not to ontologies developed with an unknown maintenance model. Many ontologies were developed by students as part of their dissertation or projects and are not maintained anymore. While this is often no issue for us as a research community, it causes problems for the industry. Some industry and government branches, for instance, are only allowed to use technologies and specifications that have been standardized before. Similarly, if you develop an atomic ontology design pattern, you may not want to align it.

That’s true. LOV is not pretending to list all the vocabularies out there. However the main philosophy would be that: if you thing your vocabulary can be reused by others, just make it visible in LOV. I do understand that companies might decide not to publish their vocab, but the community can be proud to show them some samples of ontologies ready to be reused for free in LOV. This is also part of the dissemination of SemWeb stack layer, at least the one related to ontologies. 
Regarding ODP, I guess you always reuse at least RDFS [1] or OWL [2] that are already in LOV. So I don’t see the point with the alignment here. Of course if you don’t provide a URI/labels nor metadata, the vocab won’t be added in LOV. Here you can find one example (among others) of an ODP inserted into LOV [3] 

> Best practice is always context dependent. 

Sure! 
> 
> My main issue with such an 'edited' repository with a small set of gate keepers is that the Web and the Linked Data cloud follow the AAA(AA) slogan, i.e., "Anyone can say Anything about Any topic (Anywhere and at Any time)". For instance, there are datasets on the Linked Data cloud that contain errors, are following very diverse modeling styles, and so forth. Ideally, an open vocabulary/ontology repository would only control for spam and so forth, but not for the style or engineering philosophy of the submitted ontologies. Don't get me wrong, I would love to have more ontologies submitted to LOV and use LOV frequently, but I cannot submit some of the ontologies in which I was involved for the reasons mentioned above.


There is also the notion of “quality”. Sometimes (and for some of the reasons you mentioned regarding industry and government branches), there is a need for a high quality vocab/ontology repository. AFAIK, LOV tries to play that role. Of course it does not prevent having any type of other repositories. All of them would certainly serve different purposes for the benefits of the users.


Cheers,
Ghislain 

[1] http://lov.okfn.org/dataset/lov/vocabs/rdfs <http://lov.okfn.org/dataset/lov/vocabs/rdfs>
[2] http://lov.okfn.org/dataset/lov/vocabs/owl <http://lov.okfn.org/dataset/lov/vocabs/owl>
[3] http://lov.okfn.org/dataset/lov/vocabs/odpart <http://lov.okfn.org/dataset/lov/vocabs/odpart> 

---------------------------------------
Ghislain A. Atemezing, Ph.D
Mail: ghislain.atemezing@gmail.com
Web: https://w3id.org/people/gatemezing <http://www.atemezing.org/>
Twitter: @gatemezing
About Me: https://about.me/ghislain.atemezing <https://about.me/ghislain.atemezing>