final presentation ewg-dss collab-net paper-thessaloniki
TRANSCRIPT
EWG-DSS Collab-Net
A Social Network Perspective of
DSS-Research Collaboration in Europe
EWG-DSS Collab-NetEWG-DSS Collab-Net
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-NetEWG-DSS Collab-Net
A Social Network Analysis for the EWG-DSS. An initiative of the EWG-DSS Coordination Board since 2008 (v.1)(v.1)
A joint-development involving: ALLALL the EWG-DSS Group Members and External Researchers
• Collaborators Version 2:Collaborators Version 2:Fátima Dargam, Rita Ribeiro, Pascale Zaraté, Isabelle Linden, Shaofeng Liu
David Dardenne, Alexandre Rademaker
• Collaborators Version 1:Collaborators Version 1:Fátima Dargam, Rita Ribeiro, Pascale Zaraté, Rahma Bouaziz, Tiago Simas, Shaofeng Liu
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
2
Aims of the EWG-DSSAims of the EWG-DSS
Encourage the exchange of information among Decision Systems researchers.
Facilitate international cooperation research and projects.
Promote the interest on Decision Systems in the scientific community by organizing dedicated workshops, seminars, mini-conferences, etc.
Disseminate high quality research in DSS by editing Special and Contributed Issues in relevant Scientific Journals.
Enforce the networking among its Members and among the DSS communities available.
http://ewgdss.wordpress.com 3
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
over 170 Membersover 170 Members
Ana Respício; Antonio Rodrigues; M.Eugênia Captivo; Ana Respício; Antonio Rodrigues; M.Eugênia Captivo; Tei Barnhart; Alexandre Gachet;Tei Barnhart; Alexandre Gachet; Antonio Martinez; Albert Angehrn; Alessio Ishizaka; Ana Maria Rosa Borges; Arijit BhattAntonio Martinez; Albert Angehrn; Alessio Ishizaka; Ana Maria Rosa Borges; Arijit Bhattacharya; Barbo Back; Bertrand Mareschal; Boris Delibasic ; C. Makropoulos; Jacques acharya; Barbo Back; Bertrand Mareschal; Boris Delibasic ; C. Makropoulos; Jacques Calmet; Carlos Antunes; Carlos Bana e Costa; Christer Carlsson; Asis Kr. ChattopadhCalmet; Carlos Antunes; Carlos Bana e Costa; Christer Carlsson; Asis Kr. Chattopadhyay; Caludia Loebbecke; Csaba Csaki; Dobrila Petrovic; Dorien De Tombe; Dirk Kenis; yay; Caludia Loebbecke; Csaba Csaki; Dobrila Petrovic; Dorien De Tombe; Dirk Kenis; Andreas Edelmayer ; Eduardo Natividade Jesus; Fatima Dargam; Fréféric Adam; ChristAndreas Edelmayer ; Eduardo Natividade Jesus; Fatima Dargam; Fréféric Adam; Christophe Fagot; Franck Tetard; Frits Claassen; Frieder Stolzenburg; Peter Gelleri; Gilles Coophe Fagot; Franck Tetard; Frits Claassen; Frieder Stolzenburg; Peter Gelleri; Gilles Coppin; Inès Saad; Ilya Ashikhmin; Kwakkel Jan; J. Jassbi;Tawfik Jelassi; Jeremy Forth ; Joppin; Inès Saad; Ilya Ashikhmin; Kwakkel Jan; J. Jassbi;Tawfik Jelassi; Jeremy Forth ; Jorge Souza; Joao Lourenco; Johannes Leitner; Jean Pierre Brans; Jochen Pfalzgraf; Jorgerge Souza; Joao Lourenco; Johannes Leitner; Jean Pierre Brans; Jochen Pfalzgraf; Jorge Pinho de Sousa; Jose Vincente Segura; Li Ching Ma; Lourdes Canos; Ladislav Lukas; MPinho de Sousa; Jose Vincente Segura; Li Ching Ma; Lourdes Canos; Ladislav Lukas; Marija Najika; Marko Bohannec; Philip Powel; Michael Bruhn Barfod; Miklos Biros; Mikael arija Najika; Marko Bohannec; Philip Powel; Michael Bruhn Barfod; Miklos Biros; Mikael Mihalevich; Jose Maria Moreno Jimenez; Maria Theiner; Natalio Krasnogor; Nikolaos F. Mihalevich; Jose Maria Moreno Jimenez; Maria Theiner; Natalio Krasnogor; Nikolaos F. Matsatsinis; Olaf Herden; Paul Hasenohr; Paulo Leonco; Peter Keenan; Philippe Lenca; Matsatsinis; Olaf Herden; Paul Hasenohr; Paulo Leonco; Peter Keenan; Philippe Lenca; Pierre Kunsch; Suzanne Pinson; Jean Charles Pomerol; Paulo Ramos; Rita Ribeiro; CaroPierre Kunsch; Suzanne Pinson; Jean Charles Pomerol; Paulo Ramos; Rita Ribeiro; Caroline Rieder; Rudolf Vetschera; Camille Rosenthal Sabroux; Susanne Stadler; Frantisek line Rieder; Rudolf Vetschera; Camille Rosenthal Sabroux; Susanne Stadler; Frantisek Sudzina; Maria Theiner; Thomas Soboll; Alexis Tsoukias; W. Walker; Yi Yang; Pascale Sudzina; Maria Theiner; Thomas Soboll; Alexis Tsoukias; W. Walker; Yi Yang; Pascale Zaraté; Shaofeng Liu; Jason Papathanasiou; Jorge Hernández; J. Clímaco; Dragana BecZaraté; Shaofeng Liu; Jason Papathanasiou; Jorge Hernández; J. Clímaco; Dragana Bec
4
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Main Topics of Research within the Main Topics of Research within the GroupGroup
• Collaborative Decision Making (CDM)Collaborative Decision Making (CDM) • GDSS: Group-DM; NDSS: Negotiation-DM; GDSS: Group-DM; NDSS: Negotiation-DM; • Distributed Models of Decision MakingDistributed Models of Decision Making• Applications in Collaborative Decision Making & AnalysisApplications in Collaborative Decision Making & Analysis• Models for Decision Analysis (DA) in Group Decision Making Models for Decision Analysis (DA) in Group Decision Making - Evaluation of the Collaboration Levels for DM & DA- Evaluation of the Collaboration Levels for DM & DA- Facilitation of Group Coordination and Group Communication- Facilitation of Group Coordination and Group Communication
• Knowledge Management & Context supporting Decision MakingKnowledge Management & Context supporting Decision Making • Knowledge Management as a Collaboration ModelKnowledge Management as a Collaboration Model- Knowledge-intensive Collaborative Models- Knowledge-intensive Collaborative Models- Context-based Decision Systems- Context-based Decision Systems
• Network & Web-based SystemsNetwork & Web-based Systems • Network-based Collaborative Decision MakingNetwork-based Collaborative Decision Making• New Methodologies and Technology for GDSSNew Methodologies and Technology for GDSS
• Aggregation & Fuzzy Algorithms for Decision MakingAggregation & Fuzzy Algorithms for Decision Making• Knowledge-Based (Intelligent) Decision SystemsKnowledge-Based (Intelligent) Decision Systems• Applied Decision Support Systems (including MIS)Applied Decision Support Systems (including MIS) 5
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
A Social A Social NetworkNetwork Analysis Analysis for EWG-DSSfor EWG-DSS
Motivation / Objectives:
Evaluate the group’s collaboration dynamics since its foundation (1989).
Encourage new research and promote further collaboration among its members in common projects and joint-publications.
6
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
NodesNodes: authors; publications; projects; research areas Ties / Relations:Ties / Relations: collaborations; joint-projects; Journal-editions; ...
• Distances among the members of the group.
• Major and minor areasof research concentration& interaction in the group.
• New tendencies & working areas.
• New opportunities for cooperation.
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
7
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Version 1 MethodologyVersion 1 Methodology
Weighted Graphs Methods
Software Frameworks:
NWB Network Workbench
PAJEK Network
8
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
The EWG-DSS Network V.1.0 :
Input from 70 EWG-DSS members; Period of 19 years [1989 – 2008]; 1350 Publications; 34 extracted Topics of Research Areas
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
9
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
70 Authors / Members of the EWG-DSS70 Authors / Members of the EWG-DSS
Ai Author Name Ai Author Name Ai Author Name
A1 Arijit Bhattacharya A25 Fréféric Adam A49 Pascale Zaraté
A2 Adla Abdelkader A26 Frieder Stolzenburg A50 Peter Gelleri
A3 Albert A. Angehrn A27 Frits Claassen A51 Peter Keenan
A4 Alessio Ishizaka A28 Ilya Ashikhmin A52 Philip Powel
A5 Alexis Tsoukias A29 Inès Saad A53 Philippe Lenca
A6 Ana Respício A30 J.Jassbi A54 Pierre Kunsch
A7 Antonio Jimenez Martinez A31 Jacques Calmet A55 Rita Ribeiro
A8 Asis Kr. Chattopadhyay A32 Jean Charles Pomerol A56 Rudolf Vetschera
A9 Bertrand Mareschal A33 Jean Pierre Brans A57 Sanja Petrovic
A10 Bojan Srdjevic A34 João Carlos Lourenço A58 Suzanne Pinson
A11 Boris Delibasic A35 Jochen Pfalzgraf A59 Tawfik Jelassi
A12 Caludia Loebbecke A36 Johannes Leitner A60 Thanasis Spyridakos
A13 Camille Rosenthal Sabroux A37 Jorge Freire de Sousa A61 Yi Yang
A14 Carlos Antunes A38 Jorge Pinho de Sousa A62 Thomas Soboll
A15 Carlos Bana e Costa A39 Jose Maria Moreno Jimenez A63 BAZZANA Flavio
A16 Christer Carlsson A40 Ladislav Lukas A64 Guilan Kong
A17 Csaba Csaki A41 Li Ching Ma A65 Jason Papathanasiou
A18 Dirk Kenis A42 Luís Cândido Dias A66 Mikael Mihalevich
A19 Dobrila Petrovic A43 Marko Bohannec A67 Taghezout Noria
A20 Dorien De Tombe A44 Michael Bruhn Barfod A68 Warren Elliott Walker
A21 Eduardo Manuel Natividade Jesus A45 Miklos Biros A69 José Vicente segura Heras
A22 Fatima Dargam A46 Natalio Krasnogor A70 Antonio Rodrigues
A23 Franck Tetard A47 Nguyen Dinh Pham
A24 Frantisek Sudzina A48 Olaf Herden 10
# Research Topic # Research Topic1 Business Models 18 Knowledge Management
2 Collaboration Dynamics 19 Multi-Agent Systems
3 Cooperative Decision Support Systems 20 Multiple Criteria Decision Aiding
4 Decision Analysis 21 Management Learning and Decision Making
5 Decision Aiding Process 22 Network
6 Data Mining 23 Operations research
7 Decision Support Systems 24 Preference analysis
8 Evaluation 25 Performance Evaluation
9 E-Business 26 Preference Modelling
10 Entreprise resource Planning 27 Production Planning and Scheduling
11 Expert Systems 28 Supply Chain Management
12 Economic Theory 29 Sustainable Development
13 Fuzzy Sets 30 Social Networks
14 Group Decision and Negotiation 31 Simulation Systems
15 Information Retrieval 32 Systems Software Evaluation and Selection
16 Information Systems 33 Virtual Communities
17 Information and Telecommunication Technology 34 Context
34 Topics of Research 34 Topics of Research extracted from the 1350 Publicationsextracted from the 1350 Publications::
11
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Visualization of the Publ_AP network, graphically represented in PAJEK.
Visualizationrelated by the collaboration in publications.
The graph represents how the authors’ nodes are connected among themselves, with relation to their publication-collaboration.
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
12
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Visualization of the Publ_PT represented in PAJEK, with separate components, showing the relationships among the publications and topics of research
1350 Publications
34 Topics
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
13
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
NWB Radial Graph Visualization of the Authors_AT network, showing the collaboration among the authors, with relation to their common research topics.
A65 - Jason Papathanasiou
A9 - Bertrand Mareschal
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
14
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
The collaboration relationships have shown:
1. How the members relate to each other in terms of topics of research;
2. What are the most relevant topics of research within in the group;
3. The relevant statistical data concerning our publications.
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
15
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Further analysis of this EWG-DSS Collab-Net Version 1project was also developed as a case-study of a
Master Thesis (Dardenne, 2012) from David Dardenne, supervised by Prof. Isabelle Linden from FUNDP in
Belgium, in cooperation with the EWG-DSS.
EWG-DSS Collab-Net V.1EWG-DSS Collab-Net V.1
16
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Dardenne’s Master Thesis Dardenne’s Master Thesis (Dardenne, 2012)(Dardenne, 2012)
In Dardenne’s study, the usual measures on the graph and on its nodes, as well as the measures of centrality and applications of communities detection methods were used to respond to questions like:
“Which authors were the most collaborative?”; “Among the several connected components, were there
some communities found?; and “Was there any concentration of authors in the
network?”. 17
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS – Collab-Net (Dardenne, 2012)
Intermediate ResultsIntermediate Results
Methods used:• Measures on the graph and on its nodes• Applications of communities detection methods • Measures of centrality
EURO
XXV
- Vi
lniu
s 201
2
Network R(A,P)Authors x
Publications
Relation A_APAuthors x Authors
Implementation with NodeXL
Expl. a publication with 3 authors, creation of 3 relations:
• Author 1 x Author 2• Author 1 x Author 3• Author 2 x Author 3
When 1 single author creation of 1 relation
• Author x Author
18
EWG-DSS – Collab-Net (Dardenne, 2012)
(a) Measures on the graph and nodesnodes
Network Software used: Template NodeXL
There are 73 connected components but 32 of them are single-vertex connected components; it means connected components with only one vertex
There are 782 authors
There are 3201 interactions between authors; interactions mean common publication for two authors
One of the connected components counts 2505 edges; so, we can conclude that there is a big connected component. In this connected components, 527 authors are implicated
239 publications have been written by only one author
EURO
XXV
- Vi
lniu
s 201
2
19
EWG-DSS – Collab-Net (Dardenne, 2012)
(b) CommunitiesCommunities detection Software used: Template NodeXL Graph: A(AP) Grouped by connected components Layout algorithm: Harel-Koren Fast Multiscale Type: undirected
The connected component that counts 2505 edges
EURO
XXV
- Vi
lniu
s 201
2
20
EWG-DSS – Collab-Net (Dardenne, 2012)
(b) ApplicationsApplications of communities detection methods
Software used: Template NodeXL Graph: A(AP) Grouped by clustering method GNM Layout algorithm: Harel-Koren Fast Multiscale Type: undirected
EURO
XXV
- Vi
lniu
s 201
2
21
EWG-DSS – Collab-Net (Dardenne, 2012)
(c) Measures of centrality(c) Measures of centrality
• Degree centralityNodeXL doesn’t give the degree centrality but it gives a global information on the graph about the degree distribution and gives the degree for each vertex. Here is the top 20 of the authors according to their degree level.
Software used: Template NodeXL Graph: A(AP) Control panel: “Graph metrics” ribbon ->
calculate metrics Type: undirected
EURO
XXV
- Vi
lniu
s 201
2
22
EWG-DSS – Collab-Net (Dardenne, 2012)
(c) Measures of centrality(c) Measures of centrality
• Betweeness centrality
Here is the top 20 of the authors according to their betweeness centrality level.
Software used: Template NodeXL Graph: A(AP) Control panel: “Graph metrics” ribbon -> calculate
metrics Type: undirected
EURO
XXV
- Vi
lniu
s 201
2
23
EWG-DSS – Collab-Net (Dardenne, 2012)
(c) Measures of centrality(c) Measures of centrality
• Eigenvector centrality
Here is the top 20 of the authors according to their eigenvector centrality level.
Software used: Template NodeXL Graph: A(AP) Control panel: “Graph metrics” ribbon -> calculate
metrics Type: undirected
EURO
XXV
- Vi
lniu
s 201
2
24
Dardenne’s Master Thesis Dardenne’s Master Thesis (Dardenne, 2012)(Dardenne, 2012)
Dardenne has introduced in his study the relation A x A (Authors x Authors), in which authors are linked by their common publications.
This way, the represented network could identify 782782 authors out of the original 7070 authors and members of the EWG-DSS, who contributed with the 1350 publications to start up this project.
Dardenne’s analysis has brought us one step further Dardenne’s analysis has brought us one step further on reaching our main goal.on reaching our main goal.
25
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS-Collab-Net V.2 extends the original implementation in many ways. It will consider:
1) a hybrid (manual and automatic) methodology of input data collection,using also web mining of electronic databases to automatically detect relationships of members;
2) a refined model of the publication relationship structure, taking into account “author-title-journal/conference-multiple keywords-multiple topics”;
3) as well as a more refined model of the collaboration relationship structure, which includes workshop/conference publications, informal work meetings, event co-organisations, scientific committees/boards, book/journal editorials, etc.
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
26
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS-Collab-Net V.2 collaboration analysis:
co-authorships and co-citations to further illustrate the dynamics of EWG-DSS publications overtime.
The analysis features, among other characteristics: (a) the number and percentage of multi-author papers and
co-authors in comparison with single-author papers;
(b) number and percentage of co-citations;
(c) identification of publications that are closely related to a given topic, as well as the authors involved.
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
27
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
28
Data Input
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2 ImplementationImplementation
Data Validation
Data Structure Model
Network Repository
Network AnalysisVisualisation
Web-InterfaceDissemination
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
hybrid methodology of input data collection (manual and automatic), including web mining of publications electronic databases:
DBLP Computer Science Bibliography; Academic Google; Google Scholar; Microsoft Academic Search; Private Publications URL; etc…
29
Data Input
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
Data Validation will take into account the various scripts and crawlers codes to capture and filter the relevant input information from the chosen input web-environments.
It will also cater for the validation of the publications input data (including knowledge areas, keywords) and authors’ information, as well as for its normalization.
30
Data Validation
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
31
Data InputData Input
Data ValidationData Validation : Scripts and Crawlers capture and filter Input Information
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2Data Structure Model - Network RepositoryData Structure Model - Network Repository
30 % Authors 20 % Authors 20 % Authors 20 % Authors 5 % Authors 5 % Authors
100 % Authors’ Publications validated and normalized
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
32
Considerations about the structure of the data
OntologiesWhich ready-made ontologies to adopt?bibo bibo (Bibliographic Ontology Specification)(Bibliographic Ontology Specification) ; ; foaf foaf (FOAF Vocabulary Specification)(FOAF Vocabulary Specification);; owl owl (OWL Web Ontology Language )(OWL Web Ontology Language ); ; skos skos (SKOS Core Vocabulary Specification)(SKOS Core Vocabulary Specification);; … …
Which Network Structure shall we use? Why?
Data Structure ModelNetwork Repository
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Data Model Data Model
33
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
34
Social Network Environments for Version 2(Analysis Metrics - relevant for our NetworkAnalysis Metrics - relevant for our Network)
Table with the comparison of the Analysis Functions of the Network Environments used in the EWG-DSS Collab-Net up to now.
Source: (Dardenne, 2012)
Network AnalysisVisualisation
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
35
Network Visualisation for users (EWG-DSS members)
Table with the comparison of the Visualisation Functions of the Network Environments used in the EWG-DSS Collab-Net up to now.
Source: (Dardenne, 2012)
Web-InterfaceDissemination
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Concluding RemarksConcluding Remarks
What we still have to do:
• Include missing input data (up to the current date);
• Encourage the isolated nodes of absent connections to become, at a first stage, nodes of “weak connections” within the net;
• Reduce / eliminate the isolated nodes;
• Make it available on the Internet for the use of the EWG-DSS Members;
• Bring the external collaborators (co-authors) to the EWG-DSS.36
EWG-DSS Collab-Net V.2EWG-DSS Collab-Net V.2
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
What do we need?What do we need?
To proceed we need YOUR support!To proceed we need YOUR support!
• ParticipationParticipation
• Data: Research production in joint-workData: Research production in joint-work
• Support with development forceSupport with development force
• FeedbackFeedback
37
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe
Thanks for your attention!
38
EWG-DSS Thessaloniki-2013 A Social Network Perspective of DSS-Research Collaboration in Europe