semantic and social (intra)webs
fabien gandon, @fabien_gandon, http://fabien.info
leader of the Wimmics research team (INR...
eb- nstrumented
an- achine nteractions,
ommunities, and emantics
a joint research team between INRIA Sophia Antipolis – Mé...
membersHead (and INRIA contact): Fabien Gandon
Vice Head (and I3S contact): Catherine Faron-Zucker
Researchers:
1. Michel ...
research problem
socio-semantic networks: combining formal
semantics and social semantics on the web
web landscape and graphs
(meta)data of the relations and the resources of the web
…sites …social …of data …of services
+ +...
challengesanalyzing, modeling, formalizing and
implementing graph-based social semantic web
applications for communities
...
diffusion
Master IFI: from KIS to Web
– gradual changes to the courses
– then replace the master by a new one
Standardizat...
CORESE/ KGRAM
projectsisicil.inria.fr (ANR)
 enterprise social networking
 business intelligence, watching, monitoring
 communities o...
ISICILsemantic and social intraweb for
corporate intelligence and watch
ANR project CONTINT 2009-2011
<background_knowledge>
based on Clay Shirky, TED 2009
1↔1
1→n 1→n
1→n
read-write web
internet n↔n
classical web 1↔n
wiki, (µ)blog, forum, etc. n↔n
Ward Cunningham, 94
social web networks
social networks
networking is not that new e.g. commerce
social network analysis
beginning of the 20th century
collaboratively create
and manage tags to
annotate and
categorize content
SOCIAL TAGGING
a crowd of users creating massive
categorizations
U
T
R
semantic web
mentioned by Tim BL
in 1994 at WWW
[Tim Berners-Lee 1994, http://www.w3.org/Talks/WWW94Tim/]
W3C®SEMANTIC WEB STACK
W3C®
A WEB OF
LINKED DATA
RDFstands for
Resource: pages, images, videos, ...
everything that can have a URI
Description: attributes, features, and
r...
RDFis a triple model i.e. every
piece of knowledge is broken down into
( subject , predicate , object )
doc.html has for author Fabien
and has for theme Music
doc.html has for author Fabien
doc.html has for theme Music
( doc.html , author , Fabien )
( doc.html , theme , Music )
( subject , predicate , object )
RDFtriples can be seen as arcs
of a graph (vertex,edge,vertex)
Fabien
author
doc.html
theme
Music
http://ns.inria.fr/fabien.gandon#me
http://inria.fr/schema#author
http://inria.fr/rr/doc.html
http://inria.fr/schema#theme...
principles
 use RDF as data format
 use URIs as names for things
 use HTTP URIs so that people can look up those names
...
May 2007 April 2008 September 2008
March 2009
September 2010
Linking Open Data
Linking Open Data cloud diagram, by Richard...
query with SPARQL
SPARQL Protocol and RDF
Query Language
e.g. DBpedia
->
Ontology ontologies
W3C®
PUBLISHED
SEMANTICS
OF SCHEMAS
RDFS to declare classes of
resources, properties, and
organize their hierarchy
Document
Report
creator
author
Document Per...
OWLin one…
enumeration
intersection
union
complement
 disjunction
restriction!
cardinality
1..1
algebraic properties
equi...
SKOS
knowledge
thesauri,
classifications,
subjects,
taxonomies,
folksonomies,
... controlled
vocabulary
40
natural language expressions to refer to concepts
41
inria:CorporateSemanticWeb
skos:prefLabel "corporate semantic web"@en...
between conceptsinria:CorporateSemanticWeb
skos:broader w3c:SemanticWeb;
skos:narrower inria:CorporateSemanticWiki;
skos:r...
</background_knowledge>
Social web
Semantic web
Linked data
Social semantic web
ISICIL: semantic social intraweb
isicil.inria.fr
• enterprise social networking
• business intelligence, watching, monitoring
• communities of interest, of...
proposed overview…
integrating requirement analysis methods
examples of challenges and derived functionalities
overview...
MERGING METHODOLOGIES
extracts of the requirement analysis and specifications
 analyze and model key business processes
 Analyze interactions between members of the group
ADEME « roadmap for urban m...
ex. study of transformations: existing  target
convergence matrix
detections of needs or redundancies in key scenarios
Etapes des scénarios Fonctionnalités identifiées F...
business intelligence market analysis
proposing functionalities & prototypes
Prioritization of functionalitiesFrequent functionalities and
dependencies
EXAMPLES OF FONCTIONNALITIES
examples of challenges and implementations (web 3.0 & enterprise 3.0)
assited structuring of folksonomies
flat folksonomiesweb 2.0
[Limpens et al.]
pollution
soil pollution
has narrower
pollut...
global giant graph
link users, actions, knowledge, resources, groups, etc.
#Freddy
#bk81
hasBookmark
hasTag
#tag27
industr...
pollution
pollutantpollution
pollutionpollutionpollutionpollution Soil pollutions
edition distances
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
Close match Aire sous f cm_A
Tag1 broader than Tag2 Aire sous f t1bt2_A
Related Ai...
determine thesaurus relations
Comparison of the mean value
of the JaroWinkler metric for
each type of semantic relation
Me...
Node size ↔ InDegree
◉ tags (delicious + thesenet)
◉ svic keywords
examples of results
tag1 tag2 tag3
tag1 freq (tag1) cooc (tag1, tag2) cooc (tag1, tag3)
tag2 cooc (tag2, tag1) freq (tag2) cooc (tag2, tag3)
t...
example of resultsCADIC, ADEME
environment
interest comunity inclusion
detecting narrower tags [Mika et al.]
agriculture
     12 ,,21
tagrhasnarrow...
Arrows mean "has broader"
thickness ≈ weight
examples of resultsdel.icio.us
combining metrics
edition distances
Monge-Elkan Soundex, JaroWinkler,
asymmetry Monge-Elkan Qgram
contextual metric
cosinu...
83 027 relations / 9 037 tags
 68 633 related
 11 254 hyponyms
 3 193 spelling variants
structuring as a side effect
handling conflicts
arbitration rules
IF num(narrower)/num(broader) ≥ c
THEN narrower/broader
ELSE related
purely automatic...
folksonomy enrichment lifecycle
ADDING TAGS
Automatic
processing
User-centric
structuring
Detect
conflicts
Global
structur...
Graphs, graphs, graphs
Fabien
owner
author
Researcher
type
doc.html
author
Semantic web is not antisocial
Adult
Researcher...
parentsibling
motherfatherbrothersister
colleague
knows
Gérard
Fabien
Mylène
Michel
Yvonne
<family> (guillaume)=5d(guillau...
eg. typed proximity centrality
select distinct ?y ?to
pathLength($path) as ?length
(1/sum(?length)) as ?centrality
where{
...
e.g.
ipernity.com dataset in RDF
61 937 actors & 494 510 relationships
–18 771 family links between 8 047 actors
–136 311 frien...
some interpretations
validated with managers of ipernity.com
friendOf, favorite, message, comment
small diameter, high de...
some interpretations
existence of a largest component in all sub networks
"the effectiveness of the social network at doin...
e.g. of results: different
key actors for different
kinds of links
c.f. [Erétéo et al.]
PERFORMANCES & LIMITS
Knows 0.71 s 494 510
Favorite 0.64 s 339 428
Friend 0.31 s 136 311
Family 0.03 s 18 771
Message 1.98...
SEMSNA SCHEMA
annotating the networks with analysis results
high centrality
SemSNA an ontology of SNA
http://ns.inria.fr/semsna/
4
Philippe
colleague
2
colleague
supervisor
Degree
Guillaume
Gérard
Fabien
Mylène
Michel
Yvonne
IvanPeter
example of SemSNA
ADD {
?y semsna:hasInDegree _:b0
_:b0 semsna:forProperty param[type]
_:b0 rdf:value ?indegree
_:b0 semsna:hasLength param[...
conceptual and software framework
for a semantic analysis of social networks using
semantic web frameworks
groups&reasons
detecting AND labeling communities
?
?
propagating tags
to discover communities of interest
tags to detect and label communities
extension of algorithm RAK/LP :
from random labels to structured tags
salt, water
pep...
applied to Ademe Ph.D. network
 1 853 agents
1 597 academic supervisors
256 ADEME engineers.
 13 982 relationships
10 24...
-1,2
-1,1
-1
-0,9
-0,8
-0,7
-0,6
-0,5
-0,4
-0,3
-0,2
-0,1
0
0,1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
R AK
...
results
1. pollution
2. sustainable
development
3. energy
4. chemistry
5. air pollution
6. metals
7. biomass
8. wastes
92
controled abstraction and merge
lead
PLATFORM AND PROTOTYPES
overview of the architecture
a web 3.0 solution
open integration and standard in the front-end
towardsrichwebmarks
navigating in the expert network
activity flow and notification
Semantic Web plugin for Gephi
web-scraping: archiving and integrating
Fresnel lenses to adapt results
create dynamic reports in the wiki
integrating with new web platforms
going mobile
to know more
deliverables and publications
http://isicil.inria.fr
open source code on INRIA forge
https://gforge.inria.f...
Xxxxx
xxx
Xxxxx
xxxx
xxxx
xxxxx
xxxx
Xxxxx
x
xxxxx
gave birth to …
« Unveil the semantic imprint from within your Communit...
doggy-bag
of the talk
web 1, 2
price convert?
person homepage?
more info?
web 1, 2, 3
wikipedia editions by users…
0
500000
1000000
1500000
2000000
2500000
3000000
3500000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ...
semantic spreading activation
some other contributions
mappings
linguistic patterns
[Person] is starring in [Movie]
[Perso...
informal
formal
usage representation
one web…
data
person document
program
metadata
tomorrow, he who controls the metadata,
controls the web.
@fabien_gandon
http://fabien.info
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
semantic and social (intra)webs
Upcoming SlideShare
Loading in...5
×

semantic and social (intra)webs

1,816

Published on

ICEIS 2013 Keynote talk on semantic and social (intra)webs

Published in: Technology, Education
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
1,816
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
49
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Transcript of "semantic and social (intra)webs"

  1. 1. semantic and social (intra)webs fabien gandon, @fabien_gandon, http://fabien.info leader of the Wimmics research team (INRIA, CNRS, Univ. Nice) W3C Advisory Committee Representative for INRIA
  2. 2. eb- nstrumented an- achine nteractions, ommunities, and emantics a joint research team between INRIA Sophia Antipolis – Méditerranée and I3S (CNRS and University Nice Sophia Antipolis).
  3. 3. membersHead (and INRIA contact): Fabien Gandon Vice Head (and I3S contact): Catherine Faron-Zucker Researchers: 1. Michel Buffa, MdC (UNS) 2. Olivier Corby, CR1 (INRIA) 3. Alain Giboin, CR1 (INRIA) 4. Nhan Le Thanh, Pr. (UNS) 5. Isabelle Mirbel, MdC, HDR (UNS) 6. Peter Sander, Pr. (UNS) 7. Andrea G. B. Tettamanzi, Pr. (UNS) 8. Serena Villata, RP (INRIA) Post-doc: 1. Zeina Azmeh (I3S) 2. Elena Cabrio (CORDIS) Research engineers: 1. Julien Cojan (INRIA, Ministry of Culture) 2. Christophe Desclaux (Boost your code) 3. Amosse Edouard (I3S) PhD students: 1. Pavel Arapov, 3rd year (EDSTIC-I3S) 2. Franck Berthelon, 4th year (UNS-EDSTIC) 3. Ahlem Bouchahda, 3rd year (UNS-SupCom Tunis) 4. Khalil Riad Bouzidi, 3rd year (UNS-CSTB) 5. Luca Costabello, 3rd year (INRIA-CORDI) 6. Papa Fary Diallo, 2nd year (AUF-UGB-INRIA) 7. Corentin Follenfant, 3rd year (SAP) 8. Zide Meng, 1st year (INRIA ANR OCTOPUS) 9. Maxime Lefrançois, 3rd year (EDSTIC-INRIA) 10. Nguyen Thi Hoa Hue, 2nd year (Vietnam-CROUS) 11. Nicolas Marie, 3rd year (Bell-ALU, INRIA) 12. Rakebul Hasan, 2nd year (INRIA ANR-Kolflow) 13. Oumy Seye, 2nd year, (INRIA Rose Dieng allocation) Assistants: • Christine Foggia (INRIA) • Magali Richir (I3S)
  4. 4. research problem socio-semantic networks: combining formal semantics and social semantics on the web
  5. 5. web landscape and graphs (meta)data of the relations and the resources of the web …sites …social …of data …of services + + + +… web… = + …semantics + + + +…= + typed graphs web (graphs) networks (graphs) linked data (graphs) workflows (graphs) schemas (graphs)
  6. 6. challengesanalyzing, modeling, formalizing and implementing graph-based social semantic web applications for communities  multidisciplinary approach for analyzing and modeling the many aspects of intertwined information systems communities of users and their interactions  formalizing and reasoning on these models new analysis tools and indicators new functionalities and better management
  7. 7. diffusion Master IFI: from KIS to Web – gradual changes to the courses – then replace the master by a new one Standardization participation – Working groups: RDF 1.1, SPARQL 1.1 – INRIA Advisory Committee Representative Open-source and CeCILL-C free software
  8. 8. CORESE/ KGRAM
  9. 9. projectsisicil.inria.fr (ANR)  enterprise social networking  business intelligence, watching, monitoring  communities of interest, of practice, of experts datalift.org (ANR)  from raw public data to interlinked data and schemas  a platform and documentation to assist the process  validation on real datasets kolflow.univ-nantes.fr (ANR)  reduce the overhead of communities building knowledge  federated semantic: distributed blackboard for man-machine coop. dbpedia.fr (Ministry of Culture)  extract and publish data and facts from French version of wikipedia ocktopus (ANR)  dynamic typed network of user-generated web sites (e.g. forums) DATALIFT ...
  10. 10. ISICILsemantic and social intraweb for corporate intelligence and watch ANR project CONTINT 2009-2011
  11. 11. <background_knowledge>
  12. 12. based on Clay Shirky, TED 2009 1↔1 1→n 1→n 1→n
  13. 13. read-write web internet n↔n classical web 1↔n wiki, (µ)blog, forum, etc. n↔n Ward Cunningham, 94
  14. 14. social web networks
  15. 15. social networks networking is not that new e.g. commerce social network analysis beginning of the 20th century
  16. 16. collaboratively create and manage tags to annotate and categorize content SOCIAL TAGGING
  17. 17. a crowd of users creating massive categorizations U T R
  18. 18. semantic web mentioned by Tim BL in 1994 at WWW [Tim Berners-Lee 1994, http://www.w3.org/Talks/WWW94Tim/]
  19. 19. W3C®SEMANTIC WEB STACK
  20. 20. W3C® A WEB OF LINKED DATA
  21. 21. RDFstands for Resource: pages, images, videos, ... everything that can have a URI Description: attributes, features, and relations of the resources Framework: model, languages and syntaxes for these descriptions
  22. 22. RDFis a triple model i.e. every piece of knowledge is broken down into ( subject , predicate , object )
  23. 23. doc.html has for author Fabien and has for theme Music
  24. 24. doc.html has for author Fabien doc.html has for theme Music
  25. 25. ( doc.html , author , Fabien ) ( doc.html , theme , Music ) ( subject , predicate , object )
  26. 26. RDFtriples can be seen as arcs of a graph (vertex,edge,vertex)
  27. 27. Fabien author doc.html theme Music
  28. 28. http://ns.inria.fr/fabien.gandon#me http://inria.fr/schema#author http://inria.fr/rr/doc.html http://inria.fr/schema#theme "Music"
  29. 29. principles  use RDF as data format  use URIs as names for things  use HTTP URIs so that people can look up those names  when someone looks up a URI, provide useful information (RDF, HTML, etc.) using content negotiation  include links to other URIs so that related things can be discovered
  30. 30. May 2007 April 2008 September 2008 March 2009 September 2010 Linking Open Data Linking Open Data cloud diagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/ September 2011 0 100 200 300 400 10/10/2006 28/04/2007 14/11/2007 01/06/2008 18/12/2008 06/07/2009 22/01/2010 10/08/2010 26/02/2011 14/09/2011 01/04/2012
  31. 31. query with SPARQL SPARQL Protocol and RDF Query Language
  32. 32. e.g. DBpedia
  33. 33. -> Ontology ontologies
  34. 34. W3C® PUBLISHED SEMANTICS OF SCHEMAS
  35. 35. RDFS to declare classes of resources, properties, and organize their hierarchy Document Report creator author Document Person
  36. 36. OWLin one… enumeration intersection union complement  disjunction restriction! cardinality 1..1 algebraic properties equivalence [>18] disjoint union value restrict. disjoint properties qualified cardinality 1..1 ! individual prop. neg chained prop.   keys …
  37. 37. SKOS knowledge thesauri, classifications, subjects, taxonomies, folksonomies, ... controlled vocabulary 40
  38. 38. natural language expressions to refer to concepts 41 inria:CorporateSemanticWeb skos:prefLabel "corporate semantic web"@en; skos:prefLabel "web sémantique d'entreprise"@fr; skos:altLabel "corporate SW"@en; skos:altLabel "CSW"@en; skos:hiddenLabel "web semantique d'entreprise"@fr. labels
  39. 39. between conceptsinria:CorporateSemanticWeb skos:broader w3c:SemanticWeb; skos:narrower inria:CorporateSemanticWiki; skos:related inria:KnowledgeManagement. relations
  40. 40. </background_knowledge>
  41. 41. Social web Semantic web Linked data Social semantic web ISICIL: semantic social intraweb
  42. 42. isicil.inria.fr • enterprise social networking • business intelligence, watching, monitoring • communities of interest, of practice, of experts
  43. 43. proposed overview… integrating requirement analysis methods examples of challenges and derived functionalities overview of this open-source platform http://isicil.inria.fr
  44. 44. MERGING METHODOLOGIES extracts of the requirement analysis and specifications
  45. 45.  analyze and model key business processes  Analyze interactions between members of the group ADEME « roadmap for urban mobility »  campaigns of questionnaires at Orange Labs  trend analysis of intelligence market and watch  comparison of the APIs, widgets and other applications usage analysis an specification
  46. 46. ex. study of transformations: existing  target
  47. 47. convergence matrix detections of needs or redundancies in key scenarios Etapes des scénarios Fonctionnalités identifiées Fonctions SI Présenter problématique au SVIC Mailing, Q&A Envoyer Demander ce qui est incontournable et ce que font les autres ingénieurs Consultation d’experts Extraire, filtrer Prendre en compte demande Workflow, Outils de collaboration communiquer Préparer requêtes Moteur de recherche, équation de recherche rechercher Recueillir résultats Abonnement, push… Extraire, annoter Vérifier pertinence des résultats Analyses, outils de filtrage filtrer Informer l'ingénieur Messagerie électronique, chat, vidéo-conférence envoyer S'approprier les résultats et les requêtes Equation de recherche, profil, tags Annoter, organiser Devenir le destinataire des alertes Diffusion par profil diffuser
  48. 48. business intelligence market analysis
  49. 49. proposing functionalities & prototypes Prioritization of functionalitiesFrequent functionalities and dependencies
  50. 50. EXAMPLES OF FONCTIONNALITIES examples of challenges and implementations (web 3.0 & enterprise 3.0)
  51. 51. assited structuring of folksonomies flat folksonomiesweb 2.0 [Limpens et al.] pollution soil pollution has narrower pollutant energy related related thesaurus ? SKOS
  52. 52. global giant graph link users, actions, knowledge, resources, groups, etc. #Freddy #bk81 hasBookmark hasTag #tag27 industry hasLabel #Fabien #bk34 #tag92 industries hasBookmark hasLabel hasTag
  53. 53. pollution pollutantpollution pollutionpollutionpollutionpollution Soil pollutions edition distances
  54. 54. 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Close match Aire sous f cm_A Tag1 broader than Tag2 Aire sous f t1bt2_A Related Aire sous f rel_A evaluating distancesc.f. [Limpens et al.]
  55. 55. determine thesaurus relations Comparison of the mean value of the JaroWinkler metric for each type of semantic relation Mean value of the difference s(t1,t2) - s(t2,t1) with s being the Monge-Elkan QGram metric for each set of tag pairs.
  56. 56. Node size ↔ InDegree ◉ tags (delicious + thesenet) ◉ svic keywords examples of results
  57. 57. tag1 tag2 tag3 tag1 freq (tag1) cooc (tag1, tag2) cooc (tag1, tag3) tag2 cooc (tag2, tag1) freq (tag2) cooc (tag2, tag3) tag3 cooc (tag3, tag1) cooc (tag3, tag2) freq (tag3) contextual distance: co-occurrence vector cosine distance to detect related tags   21 21 21,cos tagtag tagtag tagtag    [Cattuto et al. 2008]
  58. 58. example of resultsCADIC, ADEME
  59. 59. environment interest comunity inclusion detecting narrower tags [Mika et al.] agriculture      12 ,,21 tagrhasnarrowetaguseruser tagtag 
  60. 60. Arrows mean "has broader" thickness ≈ weight examples of resultsdel.icio.us
  61. 61. combining metrics edition distances Monge-Elkan Soundex, JaroWinkler, asymmetry Monge-Elkan Qgram contextual metric cosinus vector co-occurring tags social metrics inclusion of communities of interest football sport + +
  62. 62. 83 027 relations / 9 037 tags  68 633 related  11 254 hyponyms  3 193 spelling variants
  63. 63. structuring as a side effect
  64. 64. handling conflicts arbitration rules IF num(narrower)/num(broader) ≥ c THEN narrower/broader ELSE related purely automatic conflicting arbitrated conflict debated consensual
  65. 65. folksonomy enrichment lifecycle ADDING TAGS Automatic processing User-centric structuring Detect conflicts Global structuring Flat folksonomy Structured folksonomy [Limpens et al.]
  66. 66. Graphs, graphs, graphs Fabien owner author Researcher type doc.html author Semantic web is not antisocial Adult Researcher sub property sub class semantic web title Fabien Marco Guillaume Nicolas Michel Rémi social network analysis  ),(;)( pxrelxpdin  4)(  Guillaumedin owner Adult type
  67. 67. parentsibling motherfatherbrothersister colleague knows Gérard Fabien Mylène Michel Yvonne <family> (guillaume)=5d(guillaume)=3guillaume c.f. [Erétéo et al.]
  68. 68. eg. typed proximity centrality select distinct ?y ?to pathLength($path) as ?length (1/sum(?length)) as ?centrality where{ ?y s (foaf:knows*/rel:worksWith)::$path ?to }group by ?y      1           GEx worksWithknows c worksWithknows xkglengthkC ,/*/*
  69. 69. e.g.
  70. 70. ipernity.com dataset in RDF 61 937 actors & 494 510 relationships –18 771 family links between 8 047 actors –136 311 friend links implicating 17 441 actors –339 428 favorite links for 61 425 actors etc. c.f. [Erétéo et al.]
  71. 71. some interpretations validated with managers of ipernity.com friendOf, favorite, message, comment small diameter, high density family as expected: large diameter, low density favorite: highly centralized around Ipernity animator. friendOf, family, message, comment: power law of
  72. 72. some interpretations existence of a largest component in all sub networks "the effectiveness of the social network at doing its job" [Newman 2003] 0 10000 20000 30000 40000 50000 60000 70000 number actors size largest component knows favorite friend family message comment
  73. 73. e.g. of results: different key actors for different kinds of links c.f. [Erétéo et al.]
  74. 74. PERFORMANCES & LIMITS Knows 0.71 s 494 510 Favorite 0.64 s 339 428 Friend 0.31 s 136 311 Family 0.03 s 18 771 Message 1.98 s 795 949 Comment 9.67 s 2 874 170 Knows 20.59 s 989 020 Favorite 18.73 s 678 856 Friend 1.31 s 272 622 Family 0.42 s 37 542 Message 16.03 s 1 591 898 Comment 28.98 s 5 748 340 Shortest paths used to calculate Knows Path length <= 2: 14m 50.69s Path length <= 2: 2h 56m 34.13s Path length <= 2: 7h 19m 15.18s 100 000 1 000 000 2 000 000 Favorite Path length <= 2: 5h 33m 18.43s 2 000 000 Friend Path length <= 2: 1m 12.18 s Path length <= 2: 2m 7.98 s 1 000 000 2 000 000 Family Path length <= 2 : 27.23 s Path length <= 2 : 2m 9.73 s Path length <= 3 : 1m 10.71 s Path length <= 4 : 1m 9.06 s 1 000 000 3 681 626 1 000 000 1 000 000 )(GComp rel )(, yD rel  1 time projections
  75. 75. SEMSNA SCHEMA annotating the networks with analysis results high centrality
  76. 76. SemSNA an ontology of SNA http://ns.inria.fr/semsna/
  77. 77. 4 Philippe colleague 2 colleague supervisor Degree Guillaume Gérard Fabien Mylène Michel Yvonne IvanPeter example of SemSNA
  78. 78. ADD { ?y semsna:hasInDegree _:b0 _:b0 semsna:forProperty param[type] _:b0 rdf:value ?indegree _:b0 semsna:hasLength param[length] } SELECT ?y count(?x) as ?indegree { ?x $path ?y filter(match($path, star(param[type]))) filter(pathLength($path)<= param[length]) } group by ?y parameterized in-degree )(do lengthtype, yin  [PhD Guillaume Erétéo]
  79. 79. conceptual and software framework for a semantic analysis of social networks using semantic web frameworks
  80. 80. groups&reasons
  81. 81. detecting AND labeling communities ? ?
  82. 82. propagating tags to discover communities of interest
  83. 83. tags to detect and label communities extension of algorithm RAK/LP : from random labels to structured tags salt, water pepper, wine mustard rugby, foot foot, movie hockey sport sport sport condiment condiment condiment [Eretéo et al., 2011]
  84. 84. applied to Ademe Ph.D. network  1 853 agents 1 597 academic supervisors 256 ADEME engineers.  13 982 relationships 10 246 rel:worksWith 3 736 rel:colleagueOf  6 583 tags  3 570 skos:narrower relations between 2 785 tags
  85. 85. -1,2 -1,1 -1 -0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0 0,1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 R AK TagP S emTagP controlled S emTagP (env) controlled S emTagP (env, energ) controlled S emTagP (env, energ, model) MODULARITY COMPARISONS X axis: propagation iterations, Y axis: modularity
  86. 86. results 1. pollution 2. sustainable development 3. energy 4. chemistry 5. air pollution 6. metals 7. biomass 8. wastes 92
  87. 87. controled abstraction and merge lead
  88. 88. PLATFORM AND PROTOTYPES overview of the architecture
  89. 89. a web 3.0 solution open integration and standard in the front-end
  90. 90. towardsrichwebmarks
  91. 91. navigating in the expert network
  92. 92. activity flow and notification
  93. 93. Semantic Web plugin for Gephi
  94. 94. web-scraping: archiving and integrating
  95. 95. Fresnel lenses to adapt results
  96. 96. create dynamic reports in the wiki
  97. 97. integrating with new web platforms
  98. 98. going mobile
  99. 99. to know more deliverables and publications http://isicil.inria.fr open source code on INRIA forge https://gforge.inria.fr/projects/isicil/ models http://ns.inria.fr/
  100. 100. Xxxxx xxx Xxxxx xxxx xxxx xxxxx xxxx Xxxxx x xxxxx gave birth to … « Unveil the semantic imprint from within your Community » • cooperative society of social semantic web specialists • industrialization and maintenance of research results • developing webwarks as linked semantic traces Xxxxxx xxxxxx xxxxxxXxxxxx xxxxxx xxxxxx Xxxxxx xxxxxx xxxxxx contribute downloaded interested in exchange Xxxxxx xxxxxx xxxxxxcontributes exhange interested in contributes Xxxxxx xxxxxx xxxxxx interested in Xxxxxx xxxxxx xxxxxx link & enrich analyse & assist
  101. 101. doggy-bag of the talk
  102. 102. web 1, 2
  103. 103. price convert? person homepage? more info? web 1, 2, 3
  104. 104. wikipedia editions by users… 0 500000 1000000 1500000 2000000 2500000 3000000 3500000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 0 500000 1000000 1500000 2000000 2500000 3000000 3500000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 without bots…
  105. 105. semantic spreading activation some other contributions mappings linguistic patterns [Person] is starring in [Movie] [Person] born in [Place] (…) exploratory search natural language Q&A publication & links projection & inferences ASK { ?res creator ?prov . ?prov hasColleague ?user . ?prov interestedBy ?topic . ?user interestedBy ?topic }
  106. 106. informal formal usage representation one web… data person document program metadata
  107. 107. tomorrow, he who controls the metadata, controls the web. @fabien_gandon http://fabien.info
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×