3. > Who are disconnected
with most of people?
> Who do play the
“connector” role?
> Sub-communities?
> Central role?
Only facebook friends
Most of “my friends” are my facebook friends
Generated by TouchGraph
4. “Networks are sets of nodes
connected by edges”
Network
math
CS
physics
sociology
points lines
vertices edges, arcs
nodes links
sites bonds
actors ties, relation
= Graph
Paper
Author
Institute
Journal
Citing
Co-author
15. •G+, HIV positive gay man
•G−, HIV negative gay man
•G?, gay man, unknown HIV
•NG+, HIV positive injecting drug
using (N = needles) gay man
•NG−, HIV negative injecting drug
using gay man (n = 4);
•NG?, injecting drug using gay
man, unknown HIV status (n = 1);
•F−, HIV negative woman (n = 5);
•F?, woman, unknown HIV status (n
= 7);
•NF+, injecting drug using HIV
positive woman (n = 1);
•NF?, injecting drug using woman,
unknown HIV status (n = 1);
•M−, HIV negative heterosexual
man (n = 1); M?, heterosexual
man, unknown HIV status (n = 3);
•NM+, injecting drug using HIV
positive man (n = 1);
•NM?, injecting drug using man,
unknown HIV status (n = 1).
Potterat, et al. Risk network structure in the early epidemic phase of HIV transmission in Colorado Springs, Sexually
Transmitted Infections 78, i159–i163 (2002).
•G: gay man
•F: woman
•M: heterosexual man
•+/-/?: HIV status
•N: Injecting drug using
Contact-tracing network of sexual and injecting drug partners from 1985-1999 in Colorado Spring
16. Potterat, et al. Risk network structure in the early epidemic phase of HIV transmission in Colorado Springs,
Sexually Transmitted Infections 78, i159–i163 (2002).
Visualized by Newman, Mark EJ. "The structure and function of complex networks."
SIAM review 45.2 (2003): 167-256.
17. • Understanding the structure of sexual
networks is critical for modeling
disease transmission dynamics, if
disease is spread via sexual contact.
• This project describes the structure
of an adolescent sexual network
among a population of over 800
adolescents residing in a mid-sized
town in the mid-western US.
• Real sexual and romantic networks
are characterized by long contact
chains and few cycles.
• Implications for disease transmission
dynamics and social policy are
explored."
Data drawn from Peter S. Bearman, James Moody, and Katherine
Stovel, Chains of Affection, American Journal of Sociology 110, 44-91
(2004)
High-school dating network
18. Contact tracing network of Tuberculosis
Contagion of TB, books on politics:Valdis Krebs, www.orgnet.com.
• Black: super-spreaders
• Pink: infectious
• Green: susceptible
• Gray: uncertain
19. Role-interaction networks of movie “Les Miserables”
M. E. J. Newman and M. Girvan, Finding and evaluating community structure in networks, Physical Review E 69, 026113 (2004).
22. Who will be the most lonely one? With the
fewest social support.
Exclusion mechanism: c
a and e will spend most of their time on b and d, since each
of them only has one friend.
23. Which central role can spread rumor efficiently?
Binding mechanism: Closed one.
The central point in a closed network cannot spread rumor easily,
because its friends will double-check what it said.It is also
easier to be attacked by rumor.
24. Compared with Ke,
Lan, and Hang, Chen-
Yi shared fewer
friends with me.
However, she behaved
as a connector who
bridge several
communities of my
friendship.
Generated by TouchGraph
28. A: Depending on “what” propagate on the network
Infectious diseases
Doomsday predictionsAnti-nuclear
Health Information
Computer virusesMovie recommendations
Q: Is high efficiency(robustness) good or not?
29. • What-is and why network analysis
‣ structure’s effect rather than individual traits
• Structure’s quantitative patterns
‣ e.g., separation, clustering, degree distribution
• Why and how to become to a network
‣ Why: demographic-reasons
‣ How: considering micro-pattern: network formation
• Verify the casual relations between structure and outcome
‣ Modeling, or statistical inference
Review
30. Scale-Free: low clustered but highly-skewed degree distribution
Low degree of separation
High degree of clustering
Wang, Xiao Fan, and Guanrong Chen. "Complex networks: small-world, scale-free and beyond." Circuits and Systems Magazine, IEEE 3.1 (2003): 6-20.
31. History & Development
Freeman, Linton C. The development of social network analysis: A study in the sociology of
science. Vol. 1. Vancouver: Empirical Press, 2004.
33. • The four largest circles represent
cottages in which the girls lived.
• Each of the circles within the cottages
represents an individual girl.
• The 14 runaways are identified by
initials (e.g., HIL).
• All non-directed lines between a pair
of individuals represent feelings of
mutual attraction.
• Directed lines represent one-way
feelings of attraction.
J.#L.#Moreno,#Who#Shall#Survive?#Nervous#and#Mental#Disease#Publishing#Company,#Washington,#DC,#1934)
Moreno’s Sociogram of girls in cottages
Girls; runaway are related to their social network
34. Girls’ school dormitory dining-table partners, 1st and 2nd choices
(Moreno, The sociometry reader, 1960)
35. • Fischer(1948). Does the result of urbanization, in new cities,
destroy community? Q: What does the “community” mean?
‣ By investigating relationship between individuals, love, hate, support, and so on.
‣ 1050 adults living in 50 northern Californian communities
‣ Urbanism really reduces network density.
‣ Urbanism is negative related to psychological satisfaction and overall well-being.
• Hollingshead (1949). The adolescents’ behaviors were
strongly influenced by the “cliques” to which they belonged.
36. • Coleman(1966) Rogers (1995)
• Granovetter(1973)
Strong ties among familiar contacts increase
redundant information.
• Burt (1992)
Social capital: How contacts are connected to each other can influence
how individuals access resources
• Krackhardt (1992)
• (1981)
37. e hope to know...
• Are nodes connected? and how?
Can you connect to someone?
• How far apart are they?How many step
or how much effort do you need to get someone?
• How does a node situate on
network? Its importance
Who should you connect firstly?
• Is the network composed of
communities?How do they cluster together?
Connected
Reachable
Distance
Centrality
super-spreader
Sub-components
Belonging