About the advance towards a CASN

[hgg]

Dear SD followers,

let me give a short summary about the advance towards a CASN (Computer
Algebra Social Network) during the last two months (after the CICM-14
conference in Coimbra). Note that all this has yet to be incorporated in
detail into our wiki.

The main question in my talk
http://symbolicdata.uni-leipzig.de/Presentations/cicm-14.pdf
in Coimbra (and the real doubt expressed in Stephen Watt's question "How
do you think you will manage that?") was about DDS (as defined in the
slides) and how to manage SymbolicData to avoid such fate. I raised as
central question "How to turn passive users into active ones?" and
proposed a certain roadmap:

1) Identify passive users within our "social frame" (to adopt a notion
from knowledge engineering used also by M. Kohlhase for technical
purposes).

What's the "social frame"? SD's target is "the CA community that
consists of various Symbolic Computation subcommunities", i.e., a
"social frame" that has several "smaller social frames" (the obvious
notion "subframe" to indicate the relation to the "major frame" seems -
at a second glance - inappropriate, as discussed in the ICC session 19
at the DMV-PTM joint meeting in Poznan).

What are "passive users" within that frame? Within the SD People
database http://symbolicdata.org/Data/People/ we collected yet more
than 700 records (as lightweight RDF foaf:Person's with foaf:name and
partly sd:affiliation) of people active in a certain sense in CA.

We heavily extended the former SD people database by entries obtained
from the conference reports of CICM-x, x \in {2012,2013,2014}, and
BadBoll-14. We turned digitally available descriptions (including
references to accepted papers and talks as far as available) into a
standard RDF format, see http://symbolicdata.org/Drafts/Conferences/

Compiling those conference records we matched the authors, conference
chairs and other people information with the SD People database. Hence
you can upload these conference data into your local RDF Data store and
explore it by standard RDF technology.

The data for CICM-14 are also available in our CASN RDF store
http://symbolicdata.org/casn/ and can be explored through the CASN
SPARQL endpoint http://symbolicdata.org:8891/sparql

In particular, try the query
------------------------------------------
PREFIX sd: <http://symbolicdata.org/Data/Model#>
PREFIX dcterms: <http://purl.org/dc/terms/>
select ?pn count(?a) as ?numberOfPapers
from <http://symbolicdata.org/casn/CICM-14/Papers/>
from <http://symbolicdata.org/Data/People/>
where {
?a a sd:Reference; dcterms:creator ?p .
?p a foaf:Person; foaf:name ?pn .
} order by ?pn
------------------------------------------
to learn more about the 161 people presenting papers at CICM-14.

This kind of "author disambiguation" can be solved almost manually for a
small scientific community as our "major frame" but is a true challenge
for, e.g., the Zentralblatt. We started a joint effort with Wolfram
Sperber and compiled at http://symbolicdata.org/Data/ZBMathPeople/ a
list of (for the moment) 346 matches between SD People URIs and the
ZBMath author disambiguation system. So, if you are really listed there,
we provide (also to any interested third party) a link from your SD
person reference to Zentralblatt that can exploited by standard RDF
querying technology (unfortunately only from SD to ZBMath and not the
other direction since ZBMath does not provide RDF functionality (yet).)
It is easy to add your ZBMath reference to the system if you provide it.
The same is possible for MathReviews if we find people there similar to
Wolfram Sperber.

So much about identifying passive as potentially to activated users.

We continue to deliver information about UpcomingConferences, started a
information stack about PastConferences, provide annotated News etc. as
reported earlier, so that there is no need to go into detail about these
CASN offers here.

2) How to turn a small part of passive users identified so far into
active ones?

The easiest way is to join our mailing list, provide information and
personally benefit from a better understanding of the information and
information structure already available.

Unfortunately we did not yet succeed but on very casual responses with
similar "appeals to the common".

You (or your scientific group) can greatly benefit from our
infrastructure if you provide at a personally managed place RDF
information about your scientific activities in a consensually agreed
structured way. This is best done with a "FOAF Profile", i.e., an
extended version of the "lightweight profile" in the SD People database.
The idea is that you manage such a profile at a local site and we link
to it.

To start with we compiled such FOAF-Profiles for a limited number of
people (from the recent and former boards of the German Fachgruppe) and
stored them in the CASN RDF store. You can get a list of the profile
URIs querying

------------------------------------------
select ?p ?name
from <http://symbolicdata.org/casn/FOAF-Profiles/>
where {
?p a foaf:Person; foaf:name ?name .
} order by ?name
------------------------------------------

Such profiles are references with foaf:PersonalProfileDocuments, see
http://symbolicdata.org/casn/PersonalProfileDocuments/ for those people.
Each such profile document contains a link to the FOAF-Profile as
foaf:primaryTopic, i.e., it is easy to exchange our temporarily created
profile with yours. This profile document links also to the SD People
database via rdfs:seeAlso. Hence there is a chain of links to the "full
SD world" from your FOAF profile that can be followed up by standard RDF
tools.

3) A further step, not yet implemented, is a sioc:UserAccount with our
Facebook like semantic aware xodx-system. FOAF-profiles are also very
useful for authentication with WebID standard
http://www.w3.org/wiki/WebID using browser certificates.

So much for today about SD advances at the "people front".

Best regards, hgg

--

apl. Prof. Dr. Hans-Gert Graebe, Inst. Informatik, Univ. Leipzig
postal address: Postfach 100920, D-04009 Leipzig
Hausanschrift: Augustusplatz 10, 04109 Leipzig, Raum P-633
tel. : +49-341-97-32248
email: gra...@informatik.uni-leipzig.de
Home Page: http://www.informatik.uni-leipzig.de/~graebe