- From: Thomas Passin <list1@tompassin.net>
- Date: Tue, 22 Jul 2014 09:38:46 -0400
- To: Sarven Capadisli <info@csarven.ca>
- CC: SW-forum <semantic-web@w3.org>
On 7/22/2014 7:40 AM, Sarven Capadisli wrote: > http://csarven.ca/sense-of-lsd-analysis On reading the paper, I see you tried to do on a large scale much what I did on a small scale in 2003, namely, to get some mileage out of titles only. I found a way to enhance the semantic content of these particular titles (browser bookmarks). I found the approach to be of value in retrieval. You might like to read my paper on this, "Browser bookmark management with Topic Maps", at http://conferences.idealliance.org/extreme/html/2003/Passin01/EML2003Passin01.html I also played with a very different title clustering method, and you might like reading that work, too: "On-the-fly Clustering As A Novel RDF Query Mechanism" at http://conferences.idealliance.org/extreme/html/2004/Passin01/EML2004Passin01.html Tom Passin > Okay, so, that's boring supposedly science-magic stuff. > > Go ahead and dereference the URI to RDF. > > Once again, IMHO, what's cooler is that it is a human and > machine-friendly document. This is where Linked Research (aka: Linked > Science, Semantic Publishing etc.) comes in: > > The document is in XHTML+RDFa and has screen and print stylesheets. The > screen styles are what you would normally see in your Web browser. The > print style is based on the LNCS template (you know, the one that some > SW/LD research events force you to use when you submit your SW/LD > bling-bling in PDF) - so, yes, you can output to PDF. Go ahead and copy > the stylesheets and make it better: > http://github.com/csarven/linked-research > > See how some of the following vocabularies/ontologies are put to use: > > * Semantic Publishing and Referencing: http://purl.org/spar > * Provenance Ontology: http://www.w3.org/TR/prov-o > * Open Provenance Model for Workflows: http://www.opmw.org/ontology > * DC Terms: http://purl.org/dc/terms > > There is much room for improvement. No doubt. > > The SW/LD research community produces incredible work. Yet, what super > sucks is that the community can not get its act together to eat its own > dogfood. The community is at best stuck on *1-star* > http://www.w3.org/DesignIssues/LinkedData . Even workshops that are > about "Linked Science" or "Semantic Publishing" etc. are going in > circles. Mind-boggling. > > The community is socially challenged to improve the state of SW > research. It has a hard time learning from its own discoveries because > it is stuck on desktop native document formats e.g., PDF, as opposed to > taking it to the Web in its truest sense. It "hacks" around to attach or > gather metadata about the research document instead of focusing on the > valuable things inside those documents, which goes far beyond titles, > abstracts, subjects, references.. The community simply can not > intelligently mine previously published, *publicly funded* research. > Reinvents. The community has to jump through hoops and fire to access a > PDF document that resides in some publishers website. Whoever is in > charge of the domain/path, they call the shots! > > > Here is the challenge and a call to all SW/LD researchers. If you think > your work is interesting enough, even slightly, willing to put your neck > out, and want to make an honest contribution towards what we are all > *essentially* working on, please give this a try: > > 1. Create and publish your goods: "Any resource of significance should > be given a URI. " - > http://www.w3.org/DesignIssues/Axioms.html#Universality2 . That goes > beyond the "document" that you submit your work to conferences. It is > from hypotheses, experiments, results, workflows, everything in those > documents that deserve to be known and accessible. It is so that the > next researcher can *honestly* take from where you left off or compare > their work with yours. Don't worry, there is plenty of information that > needs to be text-mined, but we can certainly improve the situation on > what can be structured and eventually queried for. At least we have a > way to look up those atomic resources or discover them. > > 2. Publish your work on your personal site, university, work, wherever. > The point is that you should have control over it. > > 3. Link to other people's goods. > > 4. Have an open comment system policy. Get reviews, feedback, questions > all in there that the community can engage in to improve the research > further. It will feed itself. > > > That's it. I'm done. > > -Sarven > http://csarven.ca/#i > >
Received on Tuesday, 22 July 2014 13:39:13 UTC