View
911
Download
1
Category
Preview:
DESCRIPTION
Citation preview
Introduction to VIVOPaul Albert & Ryan CobineCode4LibFebruary 7, 2011
VIVO aims to address certain intractable problems of academia
- Finding collaborators - Automatically generating list of all a person’s
publications- Inferring a researcher’s expertise- Outputting a researcher’s work in a standard format
such as a CV or a biosketch.- Easily moving institutional data from one system to
another
Index lists most classes
Browse by class
Search usingfacets
Individual profile
Faculty Affairs
Scopus
User Input
Grants DB
Course management
Faculty Affairs
Co-author visualizations
Is VIVO really Facebook for researchers???
FarmVille application for VIVO currently in development!Version 1.0 drops June 2011
Opensource
Not Facebook Reason #1
Not Facebook Reason #2
Opendata
Not Facebook Reason #3
VIVO’s data is from authoritarian sources.
LOL he means authoritative #hatemyjob
Embraces semantic approach
Not Facebook Reason #4
2003 – VIVO created for local use at Cornell University (Ithaca) by life sciences
2009 – The US National Center for Research Resources (NIH) awards the VIVO Collaboration a two-year $12.2 million grant to improve VIVO
2010 Apr. – Version 1.0 released
2010 Aug. – Version 1.12010 Oct. – First VIVO conference (NYC); version 1.1.1
2011 Feb. – Version 1.2 and Harvester version 1.0
2011 Aug. – Second VIVO conference (D.C.)
A Brief History of VIVO
!"#$%&&' ($)*%#+),-!" #$%&" '(%)*" +,-(&$.." /012" 3%&-.-4$56%2" 768" 4.%9$2" :6;9"
,%<<%=-&%2"4(6%&",%(>?-2"7-&",-(?-&@A69$(*2"B..C",(%8
$(2"3$=D%"#$5%($2"B.6E%F$*D"
G6&$?2" G>=%" 'D%&2" 4(6%&" H-I$2" 7-?$<D" 3;B&$(&$C2" G-..C" 36?*.$F%>$(2" J*$..%"
36*;D$..2" K&><J%I%&*2" ,D(6?*-<D$(" L$?*.6&M2" A$F$;;%" N->&$?O
" ($)*%#+),-' ".'
/&"#)01!"369$",-&.-&"+P0PQ"%&="RS"/012",D(6?"4%(&$?2",$;6.6%"4-*$
(-2"'$((C"4(6T2"B(6&"
4(--9?2" K8C" 4>D.$(2" B..6$" 4>?DD->?$&2" ,D(6?" ,%?$2" ,D(6?U&$" ,-
M%(2" P%.(6$" #%56?2"
3%(C" B=I%(=?2" :6*%" S$(($$2" V$-(M$" G%;92" ,D(6?" G%6&
$?2" A%$" 7$?%&-2" 3%(M$%>W"
7-D&?-&2" J%(%" '($6&$?*2" 3$M%&" H%*-(($2" N%&M" H62" /%>
.%" 3%(9$?2" G%&&%D" :-(*-&2"
:%(%C%&A%>82" K.$W%&=$(" A-;9I$..2" J%(%" A>??$.." V-&E%.$E2" :%&;C
" J;D%$X$(2" #%.$"
J;D$<<.$(2" :6;D-.%?" J9%MM?2" 3%TD$I" Y$==$(2" 36;D$.$" AO" Y$&&%&*2" K.6;6%" Y
>(&$(2"
J*$<D$&"L6..6%8?O"2$0)1$1'($)*%#+),-!"'%*C"4-(&$("+0R"/012"'%56*D%,%&=(%?D$9%(2"46&"
,D$&2"JD%&?D%&",D$&2"7$&6",-)$C2"J>($?D"#$65%?6M%8
%&62"N6&M"#6&M2"A>??$.."#>D-&2"
7-&" #>&&2" /--(&68%" V-<6&%*D2" 7>.6$" G%(=$?*C2" 4(6%&" '$$?$2" :%8(%*%H$.$2" 36;%D"
H6&&$8$6$(2" :6%&.6" 3%2" A-F$(*" GO" 3;#-&%.=2" K?69/(%=D%&V-&M%Z>2" 3%(9" /(6;$2"
N>C6&"J>&2",D6&*%&"Y%&92"K.%&"L%.?D2"4(6%&"LD$$.$(2"S$&M"L>2"K&M$.%"[-??O"""3"$4%'
546""&' ".' 7%0)4)$%!" A6;D%(=" :-$." +/-&;$" /012" A6;%(=-" B?<%=%2" #%8%(6?" Y-(($?2"
36;D%$."P$M%":$M(\&O" "86%'54#)99+':%+%1#46' 2$+;,<,%!"V$(%.=" 7-C;$" +J;(6<<?"/012"
V($M"#>&.%<2" ,%*D$(6&$" #>&&2" 4(%&*" '$..$C2" /%>.%" '
6&M2" " K&M$.%"3>(($..2" 4%(F%(%"
:-F.$2" ,%(C" YD-8%?2" 36;D%$.$$&Y(68%(;D6O" " =1+6)$>,"$' ($)*%#+),-?' 5,@' A"<)+!"
A%9$?D:%M%(%Z%&" +LRJYH" /012" '(6?U" HO" G-.8$?2" ,%$(6$G->;D6&?2" V$-(M$" 7-?$<D2"
J>&6*%"4O"'->.2"H$?.6$"#O"3;0&*-?DO""=%)&&'!"#$%&&'7%0)41&'!"&&%>%!",>(U?",-.$"+L$6.."
/012"/%>."K.F$(*2"P6;*-("4(-=?9C2"3%(9"4(-&&68%&&2"K=%8",D$(6)2"Q?;%(",(>E2"#%&"
#6;96&?-&2" A6;D%(="G>2",D(6?"G>%&M2" 0*%C'.%E2"'$&&$*
D"H$$2"/$*$("36;D$.6&62"V(%;$"
36M.6-(6?62"7-D&"A>]&M2"7%?-&"J<$;.%&=2"Y(>"Y(%&2"P6&%CP%(>MD$?$2"P6(M6."L
-&MO
VIVO Collaboration
Publishers and aggregators – Elsevier, Thomson Reuters, ORCID, Collexis, Information Today, CiteSeer, ArxivOntology – Eagle-I, BIBO, FOAF, UCSFFederal agencies – OTSP, NIH, NSF, VA, USDASearch providers – Google, Bing, YahooProfessional societies – AAASSemantic web community – DERI, Tim Berners-Lee, MyExperiment, ConceptWeb, Linked DataSchools and consortia – SURA, CTSA, CIC, CBC, HubZero, FLR, dozens of individual schoolsExisting application and service providers – over 100
Collaboration & coordination
Human resources
Individuals ortheir proxies
Data aggregators and repositories
Local systemsof record
Academic affairs
Grants databases
Clinical databases
Events calendar
Credentialing DB
Course database
Sources of data
> > >
> RDF harvestSPARQL endpoint
Local data flow in VIVO
VIVO(RDF)
data ingest ontologies
(RDF)
shared as RDF
interactiveinput
local systems of
record
national sources
SubjectWalter Mondale
Data in VIVO is stored using Resource Description Framework (RDF)
Predicatehead of
ObjectTrilateral Commission
Andrew McDonald
author of
has author
research arearesearch area for
academic staff in
academic staff
Susan Riha
Mining the record: Historical evidence for…
author of has author
teaches research area for
research area
headed by
crop management
CSS 4830head of
faculty appointment in
faculty members
taught by
featured in
features person
Semantic representation of data
NYS WRI
Cornell’s supercomputers crunch weather data to help farmers manage chemicals
Earth and Atmospheric Sciences
VIVOs can connect with one another
VIVO enables authoritative data about researchers to
join the Linked Data cloud.
DT<!^^(6;D%(=O;CM%&6%9O=$^_``a^b`^.-=^.-=@=%*%?$*?c_`b`@`d@__c;-.-($=O<&M
An accessible introduction to the semantic web
Alex RockwellUniversity of Florida
AbstractVIVO provides complete information on organizational structures of institutions.Each organization object in VIVO has parent and child organizations. Startingat any particular organization, it is easy to use a simple recursion algorithm totraverse the organizations that report up to the starting point. If the startingpoint is the institution “root”, the algorithm will produce an organizational chartfor the entire organization. Using Ruby and some open source extensions, wehave developed simple software to draw pictures of organizations. We willpresent code, algorithm, commentary and sample output. All code is availableas open source at http://github.com/arockwell/vivo_org_chart/
Drawing Graphs with Labels
• Drawing graphs with 500+ nodes and corresponding labels is extremelydifficult.
• We made two major attempts to prune the graph:• Removing all non-college organizations that are direct children of UF left 300nodes remaining. The graph in the center of the poster shows these nodes.• Removing all non-college and non-department nodes from the graph left ~150nodes. These nodes are the basis for an interactive version of the graph thatincludes labels.
• We created over 100 graphs during the making of this poster.• Tweaking the settings on graph drawing programs (Graphviz and NetworkWorkBench) consumed more time than any other part of this project.
Graph of all Colleges, Departments, Centers & Institutes at UFOverview
Further Research
Copyright Information Here
Purpose• One of the goals of VIVO is to show which organizations, faculty, staff, andstudents belong to.• UF’s academic structure is highly complex and does not correspond to itsfinancial structure.• UF does not have a facility to create organizational charts. Mostorganizational charts are created by hand.
Practical Uses
Visualizing the Organization StructureThe structure is generally regular and has 4 levels:
• University of Florida (the root of the graph)• Colleges• Departments (along with some Centers and Institutes)• Centers and Institutes
Some organizations do not fit into this pattern. For example, organizations atthe college level with no sub-organizations stick out on the graph.
Finding Data Integrity ProblemsLooking at graphs generated by the program has uncovered many problemsin our data, including missing, misplaced, and duplicated records. Withoutgraphs, we might not have been able to find these inconsistencies.
We can understood organizational structure much faster by looking at thegraph rather than manually following the links from one organization toanother.
Finding all UF Organizations
We added over 100 external organizations to VIVO during CV entry of theshowcase departments. As a result, it is no longer possible to consider allentries in our database to be UF organizations. Since SPARQL cannot dorecursive queries, there was also no way to find automatically all sub-organizations at UF.
We added a rootOrganization data property to the local ontology. Thisproperty allowed us to directly mark sub-organizations as being part of UF.Solving this problem alone likely justified the time spent writing the program.
Challenges
Program Design
Extending to People
We plan to include people in graphs for a college or department, which will be particularly challenging.
• UF’s VIVO will include close to 30,000 people by the end of the grant. • We lack reliable data linking people to departments.• We need to import data to show the heads of departments.
Drawing Organizational Charts with VIVO
University of Florida
Medicine
Liberal Arts and Sciences
Engineering
Agricultural and Life Sciences
Fine Arts
Dentistry
Nursing
Education
Health and Human Performance
Veterinary Medicine
Design, Construction and Planning
Law
Pharmacy
Business Journalism
Draw organizational charts http://vivoweb.org/files/orgLast.pdf
Repurpose content into Drupal http://bit.ly/gmm8Ng
!"!#$%&$'()(*+$$%,-%.&$%/001234567$85+94$2:7$;5:<969:=9$;9:+96&$>2*?@:)+5:$A;
Get involved
- Adopt VIVO- Provide data- Develop an application- Ask questions – vivoweb.org/contact- Chat – irc.freenode.net #vivo
Recommended