SWEET used for ESIP's OSF preprints?

[brandonnodnarb]

Could SWEET be implemented as the "taxonomy" tagging resources in the/an geoscience preprint server---potentially the one hosted by OSF.io via ESIP ( see the ESIP cluster)?

Just logging it here for discussion.

[lewismc]

Hi @brandonnodnarb can you provide more information on OSF.io ? I've never used it before.

[brandonnodnarb]

Hi @lewismc -- OSF.io is the Open Science Framework; it's a project space for the...process...of science. It has tools for collaboration (including user permissions for different parts of a project), can provide a two-way handshake to github, dropbox, google docs, etc., and has mechanisms for publishing and minting DOIs, amongst other things. They have a youtube channel for introductory information, and this getting started video might also help.

The reason I mentioned this---albeit, very briefly in my initial comment---is that ESIP have been exploring setting up a preprint server for the earth sciences, i.e. arxiv.org for earth sciences domain. The Center for Open Science (COS) (the group that engineered OSF) have built in functionality to prop up pre-print servers. As such, ESIP have been in discussions with the COS about setting up said server.

My hastily subitted issue/comment was fueled by my finding out that COS support custom controlled vocabulary/thesauri for to tag resources uploaded to a domain specific pre-print server. I'm wondering if SWEET could be such a resource. This could be a fantastic use/test case!

What kind of effort would need to go in to the current version (2.4?) to make a---for now let's call it a---SKOS compliant thesauri of Earth science terminology? If there a were a group focusing more on the concepts and nomenclature of the domain (terms, synonyms, prefLabel, etc.) without concerning themselves with the full logical domain model(s), what is the scope of work that would need to be completed to get a version of SWEET to a usable level for information retrieval (precision and recall).

There are, admittedly, a lot of questions in here that need detangling. Would this be something the community would value? To further this stream of consciousness, I can envision there being different versions of SWEET, similar in may respects to how PROV is subdivided; i.e. SWEET-T (thesaurus), SWEET-O (full logical model / ontology), SWEET-odp (set of "small" ontology patterns), etc.

This is really just me thinking out loud at this point. What do others think about using (potentiall) an Earth science pre-print server as a test case for SWEET (or, a version of it)?

[narock]

+1 from me. The emerging preprint infrastructure will need, at a minimum, a thesaurus and taxonomy. Seems like SWEET could be a natural starting point for this. I'd be interested to see how the semantics could be leveraged when searching for papers via the OSF's API.

If anyone is interested in learning more about OSF and their framework, here is their demo from our first preprint cluster call: https://www.youtube.com/watch?v=xT2yyW1c6zA&feature=youtu.be&t=40s

[lewismc]

This sounds very interesting. I think that working our work on a SWEET primer would e simplified tremendously if we were to subdivide SWEET into logical constituents. I am hoping we can get further to this if we get a surge of interest further to our announcements which I'll be sending out CoB today. I encourage you to keep the 'thinking-out-loud' sentiment coming, it is exactly the type of activity we need here just now.

[lewismc]

@brandonnodnarb some documentation on modularizing ontology

https://link.springer.com/chapter/10.1007/978-3-642-24794-1_10?no-access=true

DOI: 10.1007/978-3-642-24794-1_10

[lewismc]

@brandonnodnarb do you have any suggestions about how we progress here?

[lewismc]

@narock I saw your announcement earlier today on esip-all@. Is this a good time to get a Lab's project for use of SWEET as a tagging taxonomy for EarthArXiv?

[narock]

@lewismc I think it's a bit premature right now. The EarthArXiv advisory board is still being formed and the technical infrastructure won't be in place for a few more weeks. There has been some initial discussion of taxonomies, but I don't think we have enough in place right now to write a convincing proposal. Definitely something I'd like to see in the longer term. I'd love to use SWEET in particular and add some semantics in general. We should revisit this in the future. Should we close the issue and reopen later or just leave it open?

[lewismc]

Thank you for the context @narock lets just keep this open and we can come back to it in due course. Thanks.

[brandonnodnarb]

I sent @BruceCaron a copy of the AGU Index of terms (as RDF---I believe it was a SKOS encoding) about a month ago. This should be enough to get the @EarthArXiv project/crowd started while SWEET gains momentum.

@lewismc, is v3.0 stable (enough) to start using in this capacity?

IHHO, there will (ideally) be many vocabularies used by @EarthArXiv so users can use whichever terms they are familiar.

[lewismc]

@lewismc, is v3.0 stable (enough) to start using in this capacity?

Yes it is. All URI's resolve and the URI structure will not the changing.

[pbuttigieg]

ENVO is being used by a number of journals and other content publishing systems to add some control and sematics to the environmental keywords associated with their documents, so it's not unprecedented.

The initial adoption and use need not be complex: controlling keyword tags in certain domains and linking them to URIs. Even if such tags aren't used for semantic search indexing, etc. right away, they are at least ready.

We have pretty good communication channels open and add new ENVO content when needed, with urgency dependent on the user needs.

I imagine a similar path can be followed by SWEET. It would be a shame if we didn't coordinate ENVO/SWEET to make such adopter systems interoperable, however.

On 12 Oct 2017 12:00, "brandon" notifications@github.com wrote:

I sent @BruceCaron https://github.com/brucecaron a copy of the AGU Index of terms (as RDF---I believe it was a SKOS encoding) about a month ago. This should be enough to get the @EarthArXiv https://github.com/eartharxiv project/crowd started while SWEET gains momentum.

@lewismc https://github.com/lewismc, is v3.0 stable (enough) to start using in this capacity?

IHHO, there will (ideally) be many vocabularies used by @EarthArXiv https://github.com/eartharxiv so users can use whichever terms they are familiar.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ESIPFed/sweet/issues/17#issuecomment-336081609, or mute the thread https://github.com/notifications/unsubscribe-auth/ACK7Mo-i4Q_SVPpHJbxRHp9clnA-6SCSks5sreMxgaJpZM4NulWF .

[lewismc]

@brandonnodnarb I'm not sure that the SWEET issue tracker is the correct place for this issue anymore. If there is a code repository for OSF.io or EarthArXiv then please link to this issue. I'm going to close the issue. Please reopen if you want to discuss further.