Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick...
Transcript of Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick...
![Page 1: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/1.jpg)
![Page 2: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/2.jpg)
Integration. Nick Athanassiou and friends4:30-5:00
Sharing Network Research inProgress. Roundtables
Analyzing Network DataSteve & David
3:15-4:30BREAK3:00-3:15
Publishing Network Research.Panel discussion.
Analyzing Network Data.Steve Borgatti
2:15-3:00
Data Collection. David Krackhardt1:30-2:15LUNCH12:15-1:30
Designing Network Research. Julie Hite & Dan Brass11:00-12:15MINI-BREAK10:53:21-11:00
Survey of Network Research. Dan Brass9:37-10:53:20MINI-BREAK9:30-9:37Basic Concepts. Steve Borgatti8:15-9:30Welcome. Nick Athanassiou (organizer)8:00-8:15Registration & breakfast social7:30-8:00
SCHEDULETIME
![Page 3: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/3.jpg)
BasicSocial Network
Concepts
Handouts available atwww.analytictech.com/aomnetwork
AoM PDW @ DenverAugust, 2002
Steve Borgatti, Boston Collegewww.analytictech.com/borgatti
![Page 4: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/4.jpg)
What is a Network?
• A set of dyadic ties, all of the same type,among a set of actors
• Actors can be persons, organizations …• A tie is an instance of a social relation
Bill
Bob
Betsy
Bonnie
Betty
Biff
![Page 5: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/5.jpg)
Relations Among Persons
• Kinship– mother of, wife of
• Other role-based– boss of, teacher of– friend of
• Cognitive/perceptual– knows– aware of what they know
• Affective– likes, trusts
• Interactions– give advice, talks to,
fights with– sex / drugs with
• Affiliations– belong to same clubs– is physically near
Note: Content matters!Each relation yields a different structure & has different effects
![Page 6: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/6.jpg)
Simple Answers
Data drawn from Cross, Borgatti & Parker 2001.
Recent acquisition
Older acquisitions
Original company
HR Deptof LargeHealth CareOrganization
Who you ask for answers to straightforward questions.
![Page 7: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/7.jpg)
Problem Reformulation
Recent acquisition
Older acquisitions
Original company
Data drawn from Cross, Borgatti & Parker 2001.
Who you see to help you think through issues
![Page 8: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/8.jpg)
Marriage Ties AmongFlorentine Families
Data compiled by John Padgett
During Renaissancetimes
Exampleof a Network
![Page 9: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/9.jpg)
Relations Among Organizations
• As corporate entities– Buy from / sell to, leases
to, outsources to– Owns shares of,
subsidiary of– Joint ventures, cooperate
sales agreements,alliances
– Regulates
• Via their members– Personnel flows– Interlocking directorates– Personal friendships– Co-memberships
![Page 10: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/10.jpg)
BPSCAR
CM
ENT
GDO
HCM
HR
IM
MC
MED
MH
MSRMOC
OM
OMT
ODC
OB
OCIS
ONE
PN
RM
SIM
TIM
AoM Co-Membership
> 27% overlap
![Page 11: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/11.jpg)
Internet Alliances
AOLMicrosoft
Yahoo
AT&T
Exampleof a Network
![Page 12: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/12.jpg)
Kinds of Network Data
+**2-mode
++++++++1-modeEgoComplete
Patient
PDRMerck
manual
Web MDDr. JonesBill
MomJane
Bill
Bob
Betsy
Bonnie
Betty
Biff
![Page 13: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/13.jpg)
Data collected by Cross
1-mode complete network
Informationflow withinvirtual group
![Page 14: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/14.jpg)
1-mode ego networkCarter Administration meetings
Data courtesy of Michael LinkYear 1 Year 4
![Page 15: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/15.jpg)
Ego Network Analysis
• Combine the perspective of network analysiswith the data of mainstream social science
• No computer programs available
NetworkAnalysis
MainstreamSocial Science
EgoNetworks
perspectivedata
![Page 16: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/16.jpg)
2-mode Ego Network
Patient
PDRMerck
manual
Web MDDr. JonesBill
MomJane
![Page 17: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/17.jpg)
2-mode complete network
Data compiled from newspaper societypages by Davis, Gardner & Gardner
![Page 18: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/18.jpg)
![Page 19: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/19.jpg)
The Network Perspective
• Relations vs. Attributes– Individual characteristics only half the story– People influence each other, ideas & materials flow– Predicting adoption of innovation– Interdependence vs atomistic essentialism
• Structure vs. Composition– It’s not just the elements of a system, but how
they are put together– non-reductionist, holistic, systemic
![Page 20: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/20.jpg)
The Network Perspective
• Emergence vs. Design– groups (e.g., communities vs. departments)– roles
• Structuralism vs individualism– structure -> group performance– position -> opportunities & constraints– Faith that social capital trumps human capital– more research on consequences of network structure &
position than causes– Preference for direction of causality
• position -> personality, not the reverse
![Page 21: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/21.jpg)
Graph Theoretic Concepts
![Page 22: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/22.jpg)
Directed vs undirected ties
• Undirected relations– Attended meeting with– Communicates daily with
• Directed relations– Lent money to
• Logically vs empirically directed ties– Empirically, even un-
directed relations canbe non-symmetric due tomeasurement error Bob
Betsy
Bonnie
Betty
Biff
![Page 23: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/23.jpg)
Strength of Tie
• We can attach values to ties,representing quantitative attributes– Strength of relationship– Information capacity of tie– Rates of flow or traffic across tie– Distances between nodes– Probabilities of passing on information– Frequency of interaction
Bob
Betsy
83
Jane
61
![Page 24: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/24.jpg)
Adjacency MatricesFriendship
Jim Jill Jen JoeJim - 1 0 1Jill 1 - 1 0Jen 0 1 - 1Joe 1 0 1 -
ProximityJim Jill Jen Joe
Jim - 3 9 2Jill 3 - 1 15Jen 9 1 - 3Joe 2 15 3 -
Jim
Jill
Jen
Joe
3
2
9
1
153
![Page 25: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/25.jpg)
Data FormatsDl n = 5Format = fullmatrixLabels embeddedData:
Dl n = 5Format = edgelistLabels embeddedData:Billy jillBilly john 6.3Dick janeJim bob 2.5
Dl n = 5Format = nodelistLabels embeddedData:Billy jill john jim janeJill billy bob berthaDick janeJim bob billy brenda
(No values possible) (Values optional – assigned 1 if omitted)
0101mary1000jill0101john0010billy
maryjilljohnbilly
(Values optional)
![Page 26: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/26.jpg)
1
2
3
4 5
6
7
89
10
11
12
Walks, Trails, Paths
• Path: can’t repeat node– 1-2-3-4-5-6-7-8– Not 7-1-2-3-7-4
• Trail: can’t repeat line– 1-2-3-1-7-8– Not 7-1-2-7-1-4
• Walk: unrestricted– 1-2-3-1-2-7-1-7-1
![Page 27: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/27.jpg)
Length & Distance
• Length of a path isnumber of links
• Distance between twonodes is length ofshortest path (akageodesic)
1
2
3
4 5
6
7
89
10
11
12
![Page 28: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/28.jpg)
Geodesic Distance Matrix
0124334g
1013223f
2102112e
4320123d
3211012c
3212101b
4323210a
gfedcba
![Page 29: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/29.jpg)
Components
• Maximal sets of nodes in which every nodecan reach every other by some path (nomatter how long)
• A connected graph has just one component
Relations form different networks. Components don’t.
![Page 30: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/30.jpg)
A network with 4 components
Recent acquisition
Older acquisitions
Original company
Data drawn from Cross, Borgatti & Parker 2001.
Who you go to so that you can say ‘I ran it by ____, and she says ...’
![Page 31: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/31.jpg)
Independent Paths
• A set of paths is node-independent if they share nonodes (except beginning and end)– They are line-independent if they share no lines
ST
• 2 node-independent paths from S to T• 3 line-independent paths from S to T
![Page 32: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/32.jpg)
Cutpoints
• Nodes which, if deleted, would disconnect net
Bill
Bob
Betsy
Bonnie
Betty
Biff
![Page 33: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/33.jpg)
a
b
c
d
e
f
g
h i j
k
l
m
n
o
p
q
r
s
Bridge
• A tie that, if removed, would disconnect net
![Page 34: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/34.jpg)
Local Bridge of Degree K
• A tie that connects nodes that wouldotherwise be at least k steps apart
AB
![Page 35: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/35.jpg)
Granovetter’s SWT Theory
• Strong ties create transitivity– Two nodes connected by a strong tie will have
mutual acquaintances (ties to same 3rd parties)• Ties that are part of transitive triples cannot
be bridges or local bridges• Therefore, only weak ties can be bridges
– Hence the value of weak ties• Strong ties embedded in tight homophilous
clusters, weak ties connect to diversity– Weak ties a source of novel information
![Page 36: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/36.jpg)
Granovetter Transitivity
A
B
C
D
![Page 37: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/37.jpg)
Network Cohesion
![Page 38: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/38.jpg)
Connectivity
• Line connectivity � isthe minimum number oflines that must beremoved to disconnectnetwork
• Node connectivity κ isminimum number ofnodes that must beremoved to disconnectnetwork
ST
![Page 39: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/39.jpg)
Fragmentation
• Proportion of pairs of nodes that areunreachable from each other
)1(
)1(1
�
�
��
�
nn
ssF k
kk
![Page 40: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/40.jpg)
Density
• Number of ties, expressed as percentage of the number ofordered/unordered pairs
Low Density (25%)Avg. Dist. = 2.27
High Density (39%)Avg. Dist. = 1.76
![Page 41: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/41.jpg)
Help With the Rice Harvest
Data from Entwistle et al
Village 1
![Page 42: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/42.jpg)
Help With the Rice Harvest
Whichvillageis morelikely tosurvive?
Village 2Data from Entwistle et al
![Page 43: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/43.jpg)
Average Distance
• Average distance between all pairs of nodes
Core/Peripheryc/p fit = 0.97, avg. dist. = 1.9
Clique structurec/p fit = 0.33, avg. dist. = 2.4
![Page 44: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/44.jpg)
Centralization
• Degree to which network revolves around asingle node
Carter admin.Year 1
![Page 45: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/45.jpg)
Transitivity
• Proportion of triples with 3 ties as aproportion of triples with 2 or more ties– Aka the clustering coefficient
T
A
B C
DE
{C,T,E} is atransitive triple,but {B,C,D} is not
cc = 2/6 = 33%
![Page 46: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/46.jpg)
Core/Periphery Structures
• Does the network consist of asingle group (a core) togetherwith hangers-on (a peri-phery), or
• are there multiple sub-groups, each with their ownperipheries?
C/P struct.
Cliquestruct.
![Page 47: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/47.jpg)
Subgroup Cohesion
![Page 48: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/48.jpg)
Graph-Theoretic Concepts
• Structural definitions of groups– Clique– N-clique, n-clan, n-club– K-core, K-plexes– Ls-set, Lambda sets– Factions
![Page 49: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/49.jpg)
Clique
• Maximal set of actors in which every actors isconnected to every other
• Properties– Maximum density (1.0)– Minimum distances (avg = 1)– overlapping
a bc
d
e fa b
c
d
e fa b
c
d
e f
{c,d,e} & {b,c,d} are cliques
![Page 50: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/50.jpg)
N-Clique
• A set of nodes that are within distance n ofeach other
• Relaxes distanceaspect of cliqueconcept– 1-clique is just a
cliquea
b c
d
ef
{a,b,c,e,f} is a 2-clique
![Page 51: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/51.jpg)
K-Plex
• A set of n nodes in which every node has a tieto at least n-k others in the set– In a 1-plex, every node is connected to all but one
others in the set – i.e., is a clique
a b
c
deIs {a,b,c,d,e} a 2-plex?
{a,b,d,e} is a 2-plex:each node tied to 4-2others in set.
![Page 52: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/52.jpg)
Factions
• A set of mutually exclusivegroups of actors such thatdensity of ties within groupis greater than density ofties between groups
-010010g
1-10000f
11-0000e
000-101d
0011-10c
10011-1b
010011-a
gfedcba
Density within group: 14/18 = .78Density between groups: 4/24 = .17
![Page 53: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/53.jpg)
Individual Cohesion
![Page 54: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/54.jpg)
Centrality
• Path-based– Degree– Closeness– Betweenness– Flow betweenness– Redundancy/constraint
• Walk-based– Eigenvector– Bonacich Power– Katz– Hubbell
![Page 55: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/55.jpg)
a
b
c
d
e
f
g
h i j
k
l
m
n
o
p
q
r
s
Degree Centrality
• The number of nodes adjacent to given node
HighestDegreeCentrality
��j
iji ac
![Page 56: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/56.jpg)
a
b
c
d
e
f
g
h i j
k
l
m
n
o
p
q
r
s
Closeness Centrality
• Sum of geodesic distances to all other nodes• Inverse measure of centrality
“Highest”ClosenessCentrality
��j
iji dc
![Page 57: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/57.jpg)
a
b
c
d
e
f
g
h i j
k
l
m
n
o
p
q
r
s
HighestBetweennessCentrality
Betweenness Centrality
• Loosely: number of times that a node liesalong the shortest path between two others
��
ji ij
ikjk g
gb
,
![Page 58: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/58.jpg)
a
b
c
d
e
f
g
h i j
k
l
m
n
o
p
q
r
s
HighesteigenvectorCentrality
Eigenvector Centrality
• Iterative version of degree centrality: anode’s centrality is proportional to the sum ofcentralities of those it has ties to
![Page 59: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/59.jpg)
Structural Holes
• “cheap” betweenness
Ego Ego
Few structural holes Many structural holes: - power, info, freedom
structuralhole
![Page 60: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/60.jpg)
Structural Holes
Slide from Ron Burt
![Page 61: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/61.jpg)
Entrepreneurial Network
– sparse, flat structure– independent relations,
sustained by manager– structural holes, low
redundancy provides info &control benefits
– associated with successfulmanagers
manager
![Page 62: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/62.jpg)
Support Network
– dense, flat structure– interdependent relations
sustained by each other formanager
– few holes, high redundancycreates social support
– associated with unsuccessfulmanagers
manager
![Page 63: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/63.jpg)
Structural Similarity
![Page 64: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/64.jpg)
Network Neighborhoods
• An actor’s neighborhood is the set of actorsthey are connected to
• For directed networks:– In-neighborhood
• Actors sending ties tofocal actor
– Out-neighborhood• Set of actors receiving
ties from focal actor
a b
c
de
![Page 65: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/65.jpg)
Structural Equivalence
• Actors are structurally equivalent to theextent they have the same in-neighborhoodsand out-neighborhoods
y
r
r
w
y
a b
c d
e
Structurally equivalent nodes are colored the same
![Page 66: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/66.jpg)
Structural Equivalence
• Structurally indistinguishable– Same degree, centrality, belong to same number of
cliques, etc.– Only the label on the node can distinguish it from
those equiv to it.– Perfectly substitutable: same contacts, resources
• Face the same social environment– Similar forces affecting them
![Page 67: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/67.jpg)
Structural Equivalence
• Captures notions like niche• Location or position
– You are your friends
![Page 68: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/68.jpg)
BlockModeling
1111111111j1111111111i1111111111h1111111111g1111 111f111 111e1111 111d 111 c 111 b 111 ajihgfedcba
![Page 69: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/69.jpg)
Regular Equivalence
• B and D are structurally equivalent but what Band G?– E on left has mirror-image counterpart F
• Structural equivalence is to equality whatregular equivalence is analogy
a b
c d
e fg h
i j
![Page 70: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/70.jpg)
Regular Equivalence
• Two actors are regularly equivalent if theyare connected to equivalent others– Not necessarily same others– Not necessarily in same quantity
a b
c d
e fg h
i j
![Page 71: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/71.jpg)
Technical Definition
– For any third party i that u� i, there exists an actor jthat v � j and j is regularlyequivalent to i
– For any third party i that i� u, there exists an actor jthat j � v and j is regularlyequivalent to i
a
c
d
f
g
h
• If two actors u and v are regularlyequivalent, this implies that
![Page 72: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/72.jpg)
Regular Equivalence
• Captures notion of role counterpart– Two doctors equivalent because they have same
kinds of relations with same kinds of others, suchgiving advice to patients, giving orders to nurses,receiving products from vendors, etc.
– Works when when roles are emergent – unnamed• Captures position in hierarchies well
![Page 73: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/73.jpg)
Blockmodel View
1101100101j1101011110i1010010101h1001100011g11000 0 0 001f001 00 00 110e0001 0 0 0010d0 0 0 0 111000c0 000 001000b0 0 0 0 010000ajihgfedcba
![Page 74: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/74.jpg)
Hierarchical Positiona b
ecd
f g h i j
k
![Page 75: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/75.jpg)
Categorizing SN Concepts
CohesionDensity;avg dist;
centralization
Wholenetwork
Structural ®ular equiv
classesSimilarity
CentralityDegree;
closeness;betweenness;Struct holes
GroupsCliques;n-clique;k-plex
Connection
IndividualactorsSubsets
![Page 76: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/76.jpg)
Social Capital
• Group level concept: cohesion– pattern of ties among members of a group confers
competitive advantage• immigrant groups, organizations, countries
• Individual level concept: centrality– benefits of being well connected
• material aid• information (broadly defined)• fun, companionship, emotional support, love ...
![Page 77: Publishing Network Research - Analytic Technologies concepts 2002.pdf4:30-5:00 Integration.Nick Athanassiou and friends Sharing Network Research in Progress. Roundtables Analyzing](https://reader034.fdocuments.net/reader034/viewer/2022051723/5aba8b447f8b9a441d8bc17e/html5/thumbnails/77.jpg)
To Learn More ...
• Workshop web site:– www.analytictech.com/aomnetwork
• INSNA web site:– www.heinz.cmu.edu/project/INSNA
• Contact me via email:– [email protected]
• Papers on-line:– www.analytictech.com/borgatti/paperson.htm