W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > March 2012

Re: HCLS - visualization task force?

From: Andrea Splendiani (RRes-Roth) <andrea.splendiani@rothamsted.ac.uk>
Date: Wed, 14 Mar 2012 00:05:08 +0000
To: Matthias Samwald <matthias.samwald@meduniwien.ac.at>
CC: "<public-semweb-lifesci@w3.org>" <public-semweb-lifesci@w3.org>
Message-ID: <A95F9F6A-01A0-46C7-997C-685EA3A2BD10@rothamsted.ac.uk>

probably there is not much difference. But with Semantic Web data, insight into the structure of data is more of a key, as you don't have "schemas".
If I query a database, I would look at tables to understand a structure. If I look a at a Semantic Web knowledge base, perhaps visualization is a good way to go.

That said, I think there may be an hidden danger in relying to much on visualization: people tends to explain what they see... anyway. All example we see around of effective visualization are positive: there are also cases in which patterns highlighted mean nothing, but people still find them significant.


Il giorno 13/mar/2012, alle ore 22.47, Matthias Samwald ha scritto:

Regarding visualisation of Linked Data, it should be avoided to fall victim to what was called "the pathetic fallacy of RDF" in one publication [1], i.e., visualising RDF data as graphs because the underlying datamodel is a graph, not because the use-case at hand requires this kind of visualisation (many do not).

[1] http://eprints.ecs.soton.ac.uk/12911/1/the_pathetic_fallacy_of_rdf-33.html

I have not found many _generic_ RDF visualisations that would even be remotely useful to end-users. However, if tailored/specialised SPARQL queries are needed to distil the data for visualisation, I wonder if visualisation of data from Linked Data sources is so much different from visualizing other kinds of complex data (?)

Mattias Samwald

From: Helena Deus<mailto:helenadeus@gmail.com>
Sent: Tuesday, March 13, 2012 7:37 PM
To: Luciano, Joanne Sylvia<mailto:luciaj2@rpi.edu> ; Eric Pru'dhommeaux<mailto:eric@w3.org>
Cc: Chris Baker<mailto:denguehost@gmail.com> ; Michel Dumontier<mailto:michel.dumontier@gmail.com> ; w3c semweb hcls<mailto:public-semweb-lifesci@w3.org>
Subject: Re: HCLS - visualization task force?

Hi all,

During the telco today Eric mentioned  http://www.distilbio.com/ , which was apparently built on top of exhibit3, seems like a great way to start.

I would propose not to limit ourselves to "ways to visualize RDF graphs", but to actually try to plug linked data into data visualization paradigms already widely adopted, like dendrograms, heatmaps, sunburst plots, etc

Example of a heatmap built from a SPARQL query (medicines VS their classes):

Finally, in the same lime of thinking,  a TED talk by Jer Thorp (NYT) about visualization -http://www.youtube.com/watch?feature=player_embedded&v=Q9wcvFkWpsM

There's lots of really REALLY cool js libraries that we could use for very quickly prototyping!

On Tue, Mar 13, 2012 at 5:12 PM, Luciano, Joanne Sylvia <luciaj2@rpi.edu<mailto:luciaj2@rpi.edu>> wrote:

A task force on visualization is a great idea.  And it would make sense to have it aimed at and work with helping each of the other task forces utilize visualization capabilities.  It's something I've been thinking about with respect to the work we did on the TMO/TMKB. How do we demonstrate beyond sparql queries?


On Mar 13, 2012, at 12:58 PM, Chris Baker wrote:

> Related to this topic .....
> Changing the Equation on Scientific Data Visualization
> by: Peter Fox, James Hendler
> http://www.sciencemag.org/content/331/6018/705
> The changing face of visualisation in a world of data intensive science.
> http://eresearchau.files.wordpress.com/2011/11/peter-fox.pdf
> On Tue, Mar 13, 2012 at 1:15 PM, Michel Dumontier
> <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>> wrote:
>> Hi all,
>>   At today's HCLS Life Sciences Forum (see minutes [1]) we had Alex Garcia present on the topic of RDF visualization, indeed important for the whole of the HCLS community. In addition to a Semantic Web Journal review article [2], is anybody aware of comprehensive lists of use-cases for human computer interaction with semantic web data (e.g. browse, search, query, discovery)?
>>   Second, I would like to gauge the interest of our community in visualization, and whether it would be worthwhile to set up a task force around HCLS visualization (use cases, requirements, tools and technologies, collaborative development, deployment, evaluation). Please reply indicating your interest.
>> Cheers!
>> m.
>> [1] http://www.w3.org/2012/03/13-HCLS-minutes.html
>> [2] http://iospress.metapress.com/content/2822p340453463g1
>> --
>> Michel Dumontier
>> Associate Professor of Bioinformatics, Carleton University
>> Visiting Associate Professor, Stanford University
>> Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
>> http://dumontierlab.com<http://dumontierlab.com/>
> --
> Christopher J. O. Baker Ph. D.
> Associate Professor
> Dept. Computer Science and Applied Statistics
> University of New Brunswick, Canada
> http://ca.linkedin.com/in/christopherjobaker

Helena F. Deus
Post-Doctoral Researcher at DERI/NUIG

This message has been scanned for viruses and
dangerous content by MailScanner<http://www.mailscanner.info/>, and
we believe but do not warrant that this e-mail and any attachments thereto do not contain any viruses. However, you are fully responsible for performing any virus scanning.

This message has been scanned for viruses and
dangerous content by MailScanner, and we believe 
but do not warrant that this e-mail and any attachments
thereto do not contain any viruses. However, you are fully
responsible for performing any virus scanning.
Received on Wednesday, 14 March 2012 00:06:04 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:52:51 UTC