Noshir Contractor

1
Networks as Complex Adaptive Systems
Noshir Contractor Professor, Departments of
Speech Communication Psychology Director, Age
of Networks Initiative, Center for Advanced
Study Director, Science of Networks in
Communities - National Center for Supercomputing
Applications University of Illinois at
Urbana-Champaign nosh_at_uiuc.edu
2

• Turn on power set MODE with MODE button. You
  can confirm the MODE you chose as the red
  indicator blinks.
• Lamp blinks when (someone with) a Lovegety for
  the opposite sex set under the same MODE as yours
  comes near.
• FIND lamp blinks when (someone with) a Lovegety
  for the opposite sex set under different mode
  from yours comes near. May try the other MODES to
  GET tuned with (him/her) if you like.

3
Key Takeaways

• Networks as systems, chaotic systems, complex
  systems, self-organizing systems and complex
  adaptive systems.
• Multi-theoretical multilevel motivations for
  creating, maintaining, dissolving, and
  reconstituting network links.
• Opportunity for 3D approach to network
  Discovery, Diagnosis, Design.
• Examples Tobacco research, CI-Scope

4
Aphorisms about Networks

• Social Networks
• Its not what you know, its who you know.
• Cognitive Social Networks
• Its not who you know, its who they think you
  know.
• Knowledge Networks
• Its not who you know, its what they think you
  know.

5
Cognitive Knowledge Networks
Source Newsweek, December 2000
6
INTERACTION NETWORKS
Non Human Agent to Non Human Agent Communication
Non Human Agent (webbots, avatars, databases,
push technologies) To Human Agent
Publishing to knowledge repository
Retrieving from knowledge repository
Human Agent to Human Agent Communication
Source Contractor, 2001
7
COGNITIVE KNOWLEDGE NETWORKS
Non Human Agents Perception of Resources in a
Non Human Agent
Human Agents Perception of Provision of
Resources in a Non Human Agent
Non Human Agents Perception of what a Human
Agent knows

Human Agents Perception of What Another Human
Agent Knows
Why Tivo thinks I am gay and Amazon thinks I
am pregnant .
8
Human to Human Interactions and Perceptions
Human to Non Human Interactions and Perceptions
Non Human to Human Interactions and Perceptions
Non Human to Non Human Interactions and
Perceptions
9
Systems Complexity

• Functionalism - Goals
• Cybernetics - Feedback
• Open Systems - Environment
• Complex Systems - Rules
• Chaotic systems (from order to chaos)
• Self-organizing systems (from chaos to order)
• Complex Adaptive Systems - Fitness

10
WHY DO WE CREATE, MAINTAIN, DISSOLVE, AND
RECONSTITUTE OUR COMMUNICATION AND KNOWLEDGE
NETWORKS?
11
Social DriversWhy do we create and sustain
networks?

• Theories of self-interest
• Theories of social and resource exchange
• Theories of mutual interest and collective action

• Theories of contagion
• Theories of balance
• Theories of homophily
• Theories of proximity
• Theories of co-evolution

Sources Contractor, N. S., Wasserman, S.
Faust, K. (2006). Testing multi-theoretical
multilevel hypotheses about organizational
networks An analytic framework and empirical
example. Academy of Management Review. Monge, P.
R. Contractor, N. S. (2003). Theories of
Communication Networks. New York Oxford
University Press.
12
Structural signatures of Social Drivers
Theories of Self interest
Theories of Exchange
Theories of Balance
Theories of Collective Action
Theories of Homophily
Theories of Cognition
13
What Have We Learned About These Network
Mechanisms?

• Research typically looks at only one of these
  mechanisms
• The outcomes of these mechanisms often contradict
  one another
• Some mechanisms are studied more often than
  others
• Most research examines these mechanisms at one
  point in time
• Very limited no research investigates mechanisms
  for links in multidimensional networks (where
  nodes are people and/or digital repositories)

14
INTEGRATING MULTIPLE THEORIES AT MULTIPLE LEVELS

• Multiple complementary and contradictory theories
• Theories of Balance
• Theories of Structural Holes
• Multiple levels of analysis
• Individual Theories of Structural Holes
• Dyad Social Exchange Theory
• Triad Theories of Balance
• Subgroup Theories of Cohesion
• Global Theories of Collective Action

15
Classic Network Organizations

• 15th century Florentine families (Padgett, 1987)
• 19th century Suzuki Zaibatsu (Lynn Rao, 1995)
• Mid-20th century Keiretsu (Japan), Chaibol
  (Korea), Jituanqiye (Taiwan)
• Late-20th century Joint ventures, strategic
  alliances, consortia, collaboratories, franchises

16
Ecology of networks

• Hollywood production teams
• Construction Engineering
• Research teams
• Organizational consulting firms
• FAA initiative for free flight

17
So What ???

• R.I.P. to formal organizational structures
• Move from emergent structures to emergence of
  structures
• Focus on the mechanisms for the creation,
  maintenance, and dissolution of networks within
  an ecology of networks (network of networks)

18
Heterogeneity of Agents Relations

• Human Agents
• Institutional Agents
• Groups
• Projects
• Organizations
• Associations
• Non human agents
• Databases
• Webbots
• Avatars

• Communication
• Information retrieval
• Information allocation
• Advice
• Perception of knowledge
• Trust
• Flows
• Information
• Materials
• Money

19
A large array of theoretical mechanisms that
offer contradictory and complementary
explanations
Source Monge Contractor, 2003
20
Suggest The Need For . . .

• The study of an ecology of networks as a Complex
  Adaptive System (CAS)
• Simulations as an exploratorium to determine
  potential dynamic implications of the complex
  systems
• . or to determine the extent to which the system
  is indeterminate.

21
Examples Of Networks as CAS I

• Zeggelink models evolution of friendship networks
  based on social exchange theory, classical
  conditioning theory, social comparison theory and
  balance theory.
• Leavitt et al. developed Virtual Design Team
  (VDT) an organization model based on information
  processing theory, media richness theory,
  contingency theory, and social influence theory.

22
Examples Of Networks as CAS II

• Lin and Carley present a computational model for
  organizational performance based on contingency
  theory
• Corman offers a computational model, POWERPLAY,
  to demonstrate emergence of hierarchy based on
  structuration theory
• Contractor et al. develop a computational model
  of technology adoption based on social
  information processing and social influence
  theories.

23

• An Example
• Computational model of technology adoption based
  on social information processing theory,
  (Salancik Pfeffer, 1978), structural theory of
  action (Burt, 1982 Contractor Eisenberg,
  1990), and social influence (Fulk, Schmitz, and
  Steinfield, 1990)

24
Blanche

• A tool for specifying computational models and
  running dynamic simulations
• Models are based on objects (agents) and links
  (networks) between objects
• Theoretical mechanisms can be used to specify
  dynamic relationships between and among objects
  and links
• The results of the simulation can be
  statistically analyzed and visualized

25
ModelBuilder - specify the individual attributes
of a node, and the network links connecting them
26
Specify dynamic relationships among and between
objects and links
27
Empirical Data for Example

• Location Department of Public Works in an Army
  Fort in the South-Eastern United States
• Data for the Initial Conditions
• Communication network
• General attitude towards technology
• Technological advice network
• Self-monitoring scale
• Initial conditions for scenarios
• Scenario 1 Three different department heads
  start out using the technology
• Scenario 2 Three workers in the same department
  start out using the technology
• Research Question What is the speed and extent
  of the diffusion of technology, under the two
  different initial conditions?

28
Scenario 1Three department heads initially
communicating with the technology
29
Long term adoption network based on Scenario 1
30
Scenario 2 Three users from the same department
initially communicating with the technology
31
Long term adoption network based on Scenario 2
32
Adoption patterns for the two scenarios
33
Vision Of Networks as CAS

• A new genre of scholarship that attempts to model
  explicitly and dynamically the attributes and
  relationships among a network of agents based on
  generative mechanisms suggested by one or more
  social scientific theories.

34
What can CAS tell us about an Ecology of Networks?

• How do communication, trust, resource, and
  knowledge networks influence one another in the
  creation, maintenance, and dissolution of network
  organizations?
• How do technologies structure (and, are in turn,
  structured by) the creation, maintenance, and
  dissolution of knowledge networks among
  organizational members?

35
1. Extend theories to predict the dynamics of
a Cybercommunity (e.g., Public goods theory,
Transactive memory systems)
Iterative Refinements to theory about dynamics
of Cyberinfrastructure
Multi-level hypotheses and concepts to be measured
Generative mechanisms
Iterative Design of Cyberinfrastructure
4. Develop and introduce Cyberinfrastructure
tools to enable and metrics to evaluate
the effectiveness of a Cybercommunity (IKNOW)
3. Collect longitudinal empirical data on
activities by members of a Cybercommunity
2. Develop agent-based computational models to
assess and evaluate alternative scenarios for
the long term dynamics of a Cybercommunity (Blanc
he)
Web-based surveys and real time computer-captured
data from Cybercommunity activities and
interactions
5. Statistical methods to empirically
validate the dynamics of a Cybercommunity as
predicted by the theory and models (p and MCMC
techniques)
Model predictions of Cybercommunity
36
Integrating exogenous and endogenous processes
based on multiple theories at multiple levels
leads to many possible realizations of the network
37
Unraveling the Structural Signaturesp/Exponent
ial Random Graph Modeling (ERGM)A statistical
MRI

• The observed network is one realization of the
  many possible random realizations of the network.
• Confirmatory Network Analysis The questions of
  interest in statistical modeling is whether the
  observed network exhibits the theoretically
  hypothesized structural tendencies.

38
Empirical Illustration Co-evolution of knowledge
networks and 21st century organizational forms

• NSF KDI Initiative 1999-04. PI Noshir
  Contractor, University of Illinois.
• Co-P.I.s Bar, Fulk, Hollingshead, Monge (USC),
  Kunz, Levitt (Stanford), Carley (CMU), Wasserman
  (Indiana).
• Three dozen industry partners (global, profit,
  non-profit)
• Boeing, 3M, NASA, Fiat, U.S. Army, American Bar
  Association, European Union Project Team, Pew
  Internet Project, etc.

39
MTML analysis of information retrieval and
allocation

• Why do we create information retrieval and
  allocation links with other human or non-human
  agents (e.g., Intranets, knowledge repositories)?
• Multiple theories Transactive Memory, Public
  Goods, Social Exchange, Proximity, Contagion,
  Inertial Social Factors
• Multiple levels Actor, Dyad, Global
• UIUC Team Engineering Collaboratory David
  Brandon,Roberto Dandi, Meikuan Huang,Ed
  Palazzolo, Cataldo Dino Ruta, Vandana Singh,
  and Chunke Su)

40

• Public Goods / Transactive Memory
• Allocation to the Intranet
• Retrieval from the Intranet
• Perceived Quality and Quantity of Contribution to
  the Intranet

• Transactive Memory
• Perception of Others Knowledge
• Communication to Allocate Information

Communication to Retrieve Information

• Inertia Components
• Collaboration
• Co-authorship
• Communication

Social Exchange - Retrieval by coworkers on
other topics
Proximity -Work in the same location
41
Pulling Theories Together p/ERGM

• Using a multivariate p procedure, we combined
  the primary relations from each of the theories
  into a single analysis
• This framework allows us to test for the additive
  predictability of each theory as well as
  interaction effects between the theories
• Focus for analysis Predicting a tie between two
  actors for information retrieval based on
  multiple theories

42
Multi-theoretical p/ERGM
Theoretical Predictors of CRI
1. Social Communication 0.144 2. Perception
of Knowledge Communication to
Allocate 0.995 3. Perception of Knowledge
Provision 0.972 4. Perception of Knowledge,
Social Exchange, Social Communication 0.851
5. Perception of Knowledge, Proximity,
Social Communication 0.882
43
A contextual CAS meta-theory ofsocial drivers
for ecology of networks
44
Projects Investigating Social Drivers for
Communities
Business Applications PackEdge Community of
Practice (PG) Vodafone-Ericsson Club
for virtual supply chain management (Vodafone)
Science Applications CLEANER Collaborative
Large Engineering Analysis Network for
Environmental Research (NSF) Collaboration
for Preparedness, Response Recovery
(NSF) TSEEN Tobacco Surveillance Evaluation
Epidemiology Network (NSF, NIH, CDC)
Core Research Social Drivers for Creating
Sustaining Communities
Societal Justice Applications Cultural
Networks Assets In Immigrant Communities
(Rockefeller Program on Culture
Creativity) Economic Resilience NGO Community
(Rockefeller Program on Working Communities)
Entertainment Applications World of Warcraft
(NSF) Everquest (NSF, Sony Online
Entertainment)
45
Tobacco Surveillance, Epidemiology, and
Evaluation Network (TSEEN)

• National Cancer Institute
• Center for Disease Controls National Center for
  Health Statistics (NCHS),
• Center for Disease Controls Office of Smoking
  and Health (OSH),
• Agency for Healthcare Research and Quality
  (AHRQ),
• National Library of Medicine (NLM) and
• Non-government agencies such as the American
  Legacy Foundation.

46
Tobacco Behavioral Informatics Grid (ToBIG)
Network Referral System

• Low-tar cigarettes cause more cancer than regular
  cigarettes
• A pressing need for systems that will help the
  TSEEN members effectively connect with other
  individuals, data sets, analytic tools,
  instruments, sensors, documents, related to key
  concepts and issues

47
Network Map Example based on prototype developed
for Tobacco Surveillance Evaluation
Epidemiology Community
48
Demo of TobIG
Demo of CI-Scope
49
Key Takeaways

• Networks as systems, chaotic systems, complex
  systems, self-organizing systems and complex
  adaptive systems.
• Multi-theoretical multilevel motivations for
  creating, maintaining, dissolving, and
  reconstituting network links.
• Opportunity for 3D approach to network
  Discovery, Diagnosis, Design.
• Examples Tobacco research, CI Scope

50
SONIC Research Team Members
Nat Bulkley Postdoctoral Research Associate NCSA,
UIUC
Andy Don Research Programmer NCSA, UIUC
Steven Harper Postdoctoral Research
Associate NCSA, UIUC
Hank Green Research Scientist NCSA, UIUC
Chunke Su Graduate Research Assistant Speech
Communication, UIUC
Mengxiao Zhu Graduate Research Assistant Speech
Communication, UIUC
York Yao Research Programmer NCSA, UIUC
Diana Jimeno-Ingrum Graduate Research
Assistant Labor Industrial Relations, UIUC
Annie Wang Graduate Research Assistant Speech
Communication, UIUC
51
Acknowledgements