IM5211701 – Big Data Analytics

and Applications
Introduction to Social
Network Analysis

Instructor: Chao-Lung Yang, Ph.D.

Department of Industrial
Management
National Taiwan University of
Science and Technology

1
C.-L. Yang, Big Data, NTUST IM
 Complex Networks
 Networks = nodes + links
 Not regular, but not random
 Scale-free networks
 Universal properties

Slide Credit: Leonid E. Zhukov

2
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Financial Sector

Slide Credit: Leonid E. Zhukov

3
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Transportation

Image credit: https://www.metro.taipei/cp.aspx?n=91974F2B13D997F1

4
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Biology – Protein-Protein
Interactions

Slide Credit: Leonid E. Zhukov

5
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
Graphing the history of philosophy

Image Credit: www.coppelia.io

6
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Why Study Social Network?
 Social network = Social media + networking
 2/3 of the global internet population visit
social networks
 Nielsen, Global Faces & Networked Places, 2009
 78% of people trust the recommendation of
other consumers
 Nielsen “TRUST IN ADVERTISING” REPORT,
OCTOBER 2007
 Time spent on social networks is growing at 3X
the overall internet
 Nielsen, Global & Networked Places, 2009
7
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Social Network Analysis
 A set of relational methods for systematically
understanding and identifying connections among
actors.
 is motivated by a structural intuition based on ties
linking social actors
 is grounded in systematic empirical data
 draws heavily on graphic imagery
 relies on the use of mathematical and/or computational
models.
 Social Network Analysis embodies a range of
theories relating types of observable social spaces
and their relation to individual and group behavior.

8
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Political Blogs

Slide Credit: Leonid E. Zhukov

9
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
Social Network – Relations among
People

10
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 LinkedIn Map

Slide Credit: Yi-Lin Chen

11
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Facebook Network

Social graph 500mln people, Paul Butler, 2010

12
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 What is a network?
 Terminology
 Multiple terms mentioned are similar

Slide Credit: Leonid E. Zhukov

13
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 What is a Social Network ?
 A social structure made up of individuals (or
organizations) called "nodes", which are
tied (connected) by one or more specific
types of interdependency, such as friendship,
common interest

14
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Node vs. Links
 Actors / nodes / vertices / points
▪ Computers / Telephones
▪ Persons / Employees
▪ Companies / Business Units
▪ Articles / Books
▪ Can have properties (attributes)
 Ties / edges / arcs / lines / links
 Connect pair of actors
 Types of social relations
 Allow different kind of flows

15
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Dyads, Triads and Relations
▪ actor

▪ dyad

▪ triad

friendship
▪ relation:
▪ collection of specific ties among
kinship
members of a group
Slide Credit: Johannes Putzke
16
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Strength of a Tie
▪ Social network
▪ finite set of actors and
relation(s) defined on them
▪ depicted in graph/
sociogram
Anna, female, 27
▪ labeled graph
Ken, male, 34
▪ Strength of a Tie
▪ dichotomous vs.
valued
▪ depicted in valued
graph or signed graph
(+/-)
5 2
Slide Credit: Johannes Putzke
17
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Strength of a Tie
▪ Strength of a Tie
▪ nondirectional vs.
adjacent node to/from directional
▪ depicted in directed graphs
incident node to (digraphs)
▪ nodes connected by arcs
▪ 3 isomorphism classes
▪ null dyad
+ - ▪ mutual / reciprocal /
symmetrical dyad
▪ asymmetric /
antisymmetric dyad
▪ converse of a digraph
▪ reverse direction of all arcs
Slide Credit: Johannes Putzke
18
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Why Graph Theory ?
 Graphs used to model pair wise relations
between objects
 Generally a network can be represented by a
graph
 Many practical problems can be easily
represented in terms of graph theory

19
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph Theory - History
 Begun in 1735
 Mentioned in Leonhard Euler's paper on “Seven
Bridges of Konigsberg ” .
 Problem : Walk all 7 bridges without crossing a bridge
twice
 Eulerian graph

Slide Credit: Prem Sankar C

20
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph Theory - History
 Cycles in Polyhedra - polyhedron with no
Hamiltonian cycle

Thomas P. Kirkman William R. Hamilton

Hamiltonian cycles in Platonic graphs

Slide Credit: Prem Sankar C
21
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph Theory - History
Trees in Electric Circuits

Gustav Kirchhoff

Slide Credit: Prem Sankar C

22
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph Theory - History
Four Colors of Maps

Arthur Cayley Auguste DeMorgan

Slide Credit: Yang Mu

23
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph Definition
 A graph is a collection of nodes and edges
 Denoted by G = (V, E).
 V = nodes (vertices, points).
 E = edges (links, arcs) between pairs of nodes.
 Graph size parameters: n = |V|, m = |E|.

24
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Vertex and Edge
 Vertex /Node
 Basic Element
 Drawn as a node or a dot.
 Vertex set of G is usually denoted by V(G), or V or VG
 Edge /Arcs
 A set of two elements
 Drawn as a line connecting two vertices, called end vertices,
or endpoints.
 The edge set of G is usually denoted by E(G), or E or EG
 Neighborhood
 For any node v, the set of nodes it is connected to via an edge
is called its neighborhood and is represented as N(v)

25
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Graph: Example
 n:= 6 , m:=7
 Vertices (V) :={1,2,3,4,5,6}
 Edge (E) := {1,2},{1,5},{2,3},{2,5},{3,4},{4,5},{4,6}}
 N(4) := Neighborhood (4) ={6,5,3}

26
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Edge Types
 Undirected;
 E.g., distance between two cities, friendships…
 Directed; ordered pairs of nodes.
 Directed edges have a source (head, origin) and
target (tail, destination) vertices
 Weighted ; usually weight is associated .

27
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Empty Graph / Edgeless graph
 No edge

 Null graph
 No nodes

 Obviously no edge

28
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Simple Graph (Undirected)
 Simple Graph are undirected graphs without
loop or multiple edges
 A = AT

29
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Directed graph : (digraph)

 Edges have directions

 A !=AT

loop

multiple arc

arc node

System Informatics &

Management Science Lab
 Directed Graph (digraph)
 Edges have directions

loop

multiple arc

arc node

31
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Weighted Graph
 Weighted graph is a graph for which each
edge has an associated weight

1.2 2
1 2 3 1 2 3
.2
.5 1.5 5 3
.3 1
4 5 6 4 5 6

.5

32
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Subgraph
 Vertex and edge sets are subsets of those of
G
 a supergraph of a graph G is a graph that
contains G as a subgraph.

33
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Adjacency Matrix
 n-by-n matrix with Auv = 1 if (u, v) is an edge.
 Diagonal Entries are self-links or loops
 Symmetric matrix for undirected graphs

1 2 3 4 5 6 7 8
1 0 1 1 0 0 0 0 0
2 1 0 1 1 1 0 0 0
3 1 1 0 0 1 0 1 1
4 0 1 0 1 1 0 0 0
5 0 1 1 1 0 1 0 0
6 0 0 0 0 1 0 0 0
7 0 0 1 0 0 0 0 1
8 0 0 1 0 0 0 1 0

34
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Incidence Matrix
 VxE
 [vertex, edges] contains the edge's data
1,2 1,5 2,3 2,5 3,4 4,5 4,6

1 1 1 0 0 0 0 0
2 1 0 1 1 0 0 0
3 0 0 1 0 1 0 0
4 0 0 0 0 1 1 1
5 0 1 0 1 0 1 0
6 0 0 0 0 0 0 1

35
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Adjacency List
 Edge list  Adjacency List (node
list)
Edge List Node List
12 122
12 235
23 33
25 435
33 534
43
45
53
54

36
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
Edge Lists for Weighted Graphs

Edge List
1 2 1.2
2 4 0.2
4 5 0.3
4 1 0.5
5 4 0.5
6 3 1.5

37
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Classification of Graph Terms
 Global terms refer to a whole graph
 Local terms refer to a single node in a graph

38
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Connected and Isolated vertex
 Two vertices are connected if there is a path
between them
 Isolated vertex – not connected

1 2 3

isolated vertex 4 5 6

39
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Adjacent nodes
 Adjacent nodes
 Two nodes are adjacent if they are connected
via an edge.
 If edge e={u,v} ∈ E(G), we say that u and v are
adjacent or neighbors
 An edge where the two end vertices are the
same is called a loop, or a self-loop

40
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Degree (Undirected Graphs)
 Number of edges incident on a node

The degree of 5 is 3

41
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Degree (Directed Graphs)
 In-degree: Number of edges entering
 Out-degree: Number of edges leaving
 Degree = indeg + outdeg
outdeg(1)=2
indeg(1)=0

outdeg(2)=2
indeg(2)=2

outdeg(3)=1
indeg(3)=4

42
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Walk and Trial
 walk: a path in which edges/nodes can be
repeated.

 trail: no edge can be repeat a-b-c-d-e-b-d

43
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Paths
 Path: is a sequence P of nodes v1, v2, …, vk-1, vk
 No vertex can be repeated
 A closed path is called a cycle
 The length of a path or cycle is the number of edges visited
in the path or cycle

Walks and Paths

1,2,5,2,3,4 1,2,5,2,3,2,1 1,2,3,4,6
walk of length 5 Cycle walk of path of length 4
44
length
C.-L. Yang, Big 6 IM
Data, NTUST System Informatics &
Management Science Lab
 Cycle
 Cycle - closed path: cycle (a-b-c-d-a) ,
closed if x=y
 Cycles denoted by Ck, where k is the number
of nodes in the cycle

C3 C4 C5

45
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Walks, Trails, Paths
▪ (Directed) Walk (W)
▪ sequence of nodes and lines starting and ending with (different) nodes
(called origin and terminus)
▪ Nodes and lines can be included more than once
▪ Inverse of a (directed) walk (W-1)
▪ Walk in opposite order
▪ Length of a walk
▪ How many lines occur in the walk? (same line counts double, in
weighted graphs add line weights)
▪ (Directed) Trail
▪ Is a walk in which all lines are distinct
▪ (Directed) Path
▪ Walk in which all nodes and all lines are distinct
▪ Every path is a trail and every trail is a walk
Slide Credit: Johannes Putzke
46
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
Walks, Trails and Paths - Repetition
l3

l2 l4 l5
l7
l1
l6

▪ W = n1 l1 n2 l2 n3 l4 n5 l6 n6 ▪ Path
▪ n1 ▪ origin
▪ n3 ▪ terminus
▪ W = n1 l1 n2 l2 n3 l4 n5 l4 n3 ▪ Walk
▪ W = n1 l1 n2 l2 n3 l4 n5 l5 n4 l3 n3 ▪ Trail

Slide Credit: Johannes Putzke

47
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Shortest Path
 Shortest Path is the path between two nodes
that has the shortest length
 Length – number of edges.
 Distance between u and v is the length of a
shortest path between them
 The diameter of a graph is the length of the
longest shortest path between any pairs of
nodes in the graph

48
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
Reachability, Distances and Diameter
▪ Reachability
▪ If there is a path between
nodes ni and nj
▪ Geodesic
▪ Shortest path between two nodes
▪ (Geodesic) Distance d(i,j)
▪ Length of Geodesic (also called „degrees of separation“)

Slide Credit: Johannes Putzke

49
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Six degrees of separation
 “Any two people are on average separated
no more that by six intermediate
connections”
 John Guare play (1991) and movie (1993), ”Six
Degrees of Separation”

50
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Six degrees of separation

Slide Credit: Leonid E. Zhukov

51
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Small World

 "The small-world
problem". Stanley
Milgram, 1967
 "An experimental
study of the small
world problem J.
Travers, S.
Milgram, 1969

52
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Stanley Milgram’s 1969 experiment

 296 volunteers, 217 sent

 196 in Nebraska (1300
miles)
 100 in Boston (25 miles)
 Target in Boston
 Name, address,
occupation,
 job, hometown

53
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 54
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Result of Stanley Milgram’s 1969
experiment
 Reached the target N = 64, 29%
 Average chain length <L> = 5.2
 Channels: hometown <L> = 6.1
 Business contacts <L> = 4.6
 From Boston <L> = 4.4
 From Nebraska <l> = 5.7

55
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Small World
 Email graph:
 D. Watts (2001), 48,000 senders, <L> 6
 MSN Messenger graph:
 J. Lescovec et. al (2007), 240 million users, <L>
6.6
 Facebook graph:
 L. Backstrom et. al (2012), 721 million users (≈
69 billion friendship links), <L> 4:74

56
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab
 Cayley Tree
 Bethe lattice: infinite cycle free graph where every node
connect to z neighbors - coordination number
 per shell Nk = z(z -1)k-1
 total N = 1 + σ𝐿𝑘=1 z(z −1)k−1
 Estimates:
 zL = N, L = log N/log z
 N ≈ 6.7 billion, z = 50 friends, L ≈ 5.8

z=3
57
C.-L. Yang, Big Data, NTUST IM System Informatics &
Management Science Lab