Mining Graph Data

GRAPH THEORY REVIEW

Teacher: Le Ngoc Thanh
Email: [email protected]
Contents
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

2
Review questions
• Question 1: What components does graph G consist of?
• Question 2: What is a subgraph?
• Question 3:When are two vertices of a graph called
adjacent?
• Question 4: What is a path between two vertices?
• Question 5: What is a simple path?
• Question 6: What is a cycle?
• Question 7: What is a simple cycle?

3
Review questions
• Q: A vertex of a graph is called even or odd depending
upon?
A) Total number of edges in a graph is even or odd
B) Total number of vertices in a graph is even or odd
C) Its degree is even or odd
D) D None of these

4
Graph
• A graph G consists of two sets: a set V containing
vertices (vertices, nodes), and a set E containing
edges connecting vertices.

For example, a map of a University

is a graph with vertices being
buildings and edges being paths
between buildings.
5
Example for a graph

6
Subgraph
• The subgraph consists of a subset of the vertices and
a subset of the edges of the graph.

7
Questions
• Q1: What are the degrees of the vertices of a star?
• Q2: Let G be the graph represented in figure and X be
the set {a, b, f, e, g, l}. Draw a figure of G[X].

8
Adjacent
• Two vertices of a graph are adjacent if connected by
an edge.

The Library và the Student Union là liền kề.

9
Path
• A path between two vertices is a sequence of edges
that begins at one vertex and ends at another.

There is a route that starts from

Dormitory, through the Library,
to the Student Union, and
finally to the Gymnasium,

10
Simple Path
• A simple path is a path that does not pass through
the same vertex more than once.

11
Cycle
• A cycle is a path that begins and ends at the same
vertex; A single cycle is a cycle that does not pass
through the other vertices more than once.

Library–Student Union–
Gymnasium–Dormitory–Library
is a single cycle

12
Questions
• Draw a figure of a path of length 0, of a path of
length 1 and of a path of length 2.
• Draw a figure of a circuit of length 3 and of a circuit
of length 4.
• Why does the definition of a circuit require “n ≥ 3”?

13
Review questions
• Question 1: When is a graph called connected?
• Question 2: What is a complete graph?
• Question 3: Is a complete graph necessarily connected?
• Question 4: Can a graph have duplicate edges between
vertices?
• Question 6: Is a multigraph a graph?
• Question 7: Can the edges of a graph start and end at
the same vertex?
• Question 8: Can you label the edges of the graph?
14
Connected and Unconnected Graphs

• A graph is connected if each pair of distinct

vertices has a path between them.
• That is, in a connected graph you can go from
any vertex to any other vertex by following a
path.

15
Complete graph
• In a complete graph, each pair of distinct
vertices has an edge between them.
• A complete graph is also connected, but a
connected graph is not necessarily complete.

16
Multigraph
• Because a graph has a set of edges, a graph cannot
have duplicate edges between vertices.
• However, a multigraph allows multiple edges.
- Therefore, a multigraph is not a graph. The edges of a graph can’t
start and end at the same vertex.

17
Self edge
• The edges of a graph that start and end at the
same vertex are called self edges or loops.

18
Label
• You can label the edges of the graph.

19
Content
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

20
Review questions
• Q1: What is a weighted graph?
• Q2: What is the difference between an undirected
graph and a directed graph?
• Q3: What is a bipartite graph?
• Q4: What is a regular graph?
• Q5: Are all complete graphs regular graphs?
• Q6: What characterizes a small-world graph?
• Q7: What is a stochastic graph?

21
Weighted graph
• When the labels represent numeric values, the
graph is called a weighted graph.

The weighted graph of edges is

labeled as the distance between
cities
22
Directed and Undirected Graphs
• In an undirected graph, edges have no direction. That is, we
can move in either direction along the edges between
vertices in an undirected graph.
• Conversely, each edge in a directed graph will have a
direction and is called a directed edge.

This is a directed graph. There are flights

in both directions between Providence and
New York, but although there are flights
from San Francisco to Albuquerque, there
are no flights from Albuquerque to San
Francisco. 23
Directed and Undirected Graphs
• The definition of adjacent vertices is not exactly the same in
both types of graphs.
• In the digraph:
- If there is an edge directed from vertex x to vertex y, then y is adjacent
to x. (In other words, y is a descendant of x and x is an ancestor of y.)
And it's not necessary that x is adjacent to y .

Albuquerque borders San Francisco, but

San Francisco does not border
Albuquerque. 24
Exercises
Which of the following statements is/are TRUE for
undirected graphs?
P: Number of odd degree vertices is even.
Q: Sum of degrees of all vertices is even.
(A) Neither P nor Q
(B) Both P and Q
(C) Q only
(D) P only

25
Bipartite graph
• A bipartite graph is a graph in which vertices can be
divided into two sets that do not communicate with
each other, and all edges in the graph connect only
vertices on different sets.

26
•

Exercises
The maximum number of edges in a bipartite graph on
12 vertices is __________________________.
A) 12
B) 24
C) 36
D) 48

27
Random graph
• Random graph is the graph where the edges are
randomly generated.

28
Regular graph
• A regular graph is a graph where all vertices
have the same degree.
• From there, it can be seen that all complete graphs
are regular graphs

29
Exercises
• The given Graph is regular?

30
Scale-free graph
• A graph is said to be scale-free if its features are
independent of the size of the graph (the number of
vertices). That is, as the graph expands, the underlying
structure remains the same.

31
Small-world graph
• Small-world is a graph where the average path between
all vertices is small. That is, a vertex can reach any
vertex through only a few edges.

32
Stochastic graph
• A stochastic graph is a graph where the weights of each
edge are represented with a probability value
representing the probability of moving between two
determinists. The sum of the weights of the edges at
each vertex is 1.

33
Contents
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

34
Review questions
• Q1: What are the two most common implementations of
graphs in computer systems?
• Q2: How does an adjacency matrix represent the existence of
edges between vertices?
• Q3: What changes in an adjacency matrix when the graph is
weighted?
• Q4: How does an adjacency list represent edges in a graph?
• Q5: Which graph implementation is better, adjacency
matrices, or adjacency lists?
• Q6: How do space requirements differ between adjacency
matrices and adjacency lists?

35
Graph storage
• There are different ways to store graphs in computer
systems.
• The data structure used depends on both the graph
structure and the algorithm used to manipulate the
graph.
• The two most common implementations of graphs
are adjacency matrices and adjacency lists.

36
Adjacent Matrix
• The adjacent matrix for a graph with n vertices numbered 0, 1,
..., n - 1 is:
- An array matrix n x n such that the matrix [i] [j] is 1 if there is an edge
from vertex i to vertex j and 0 if there is none.

37
Adjacent matrix

• When the graph is weighted:

- Let the matrix [i] [j] be the side-labeled weighting from
vertex i to vertex j, instead of simply 1
- Let the matrix [i] [j] be equal to ∞ instead of 0 when there
are no sides from vertex i to vertex j.

38
 Adjacent lists
• The adjacent list for a graph with n vertices numbered 0, 1, ..., n -
1 consists of n association sequences.
- The i-link series has a node for vertex j if and only if the graph contains an
edge from vertex i to vertex j. This node may contain the value of vertex j, if
any.

39
Adjacent lists

The adjacent list for a scalar graph treats each edge as

if it were two edges with opposite directions.

40
Which way is better?
• Which of these two graph implementations—adjacent
matrices or adjacent lists—is better?
- The answer depends on how your particular application uses graphs.

• Example:
- Determine if there is an edge from vertex i to vertex j -> adjacent
matrix
- Find all vertices adjacent to a given I-vertex -> the adjacent list

41
Which way is better?
• Consider the space requirements of the two
deployments.
- On the surface, adjacent matrices require less memory than
adjacent lists, because each entry in the matrix is simply an
integer, while each list node contains both values to define
vertices and pointers.
- However, the adjacent matrix always has n2 entry, while the
number of nodes in the adjacent list is equal to the number of
edges in a directed graph or double that number for a scalar
graph. Although the adjacent list also has n leading pointers, it
usually requires less storage than the adjacent matrix

42
Question
• Do you know incidence matrix?

43
Contents
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

44
Review questions
• Q1: How can you determine the connected components
of a graph?
• Q2: Describe the difference between Depth-First Search
(DFS) and Breadth-First Search (BFS) in graph traversal.
• Q3: Define topological sorting in a directed acyclic graph.
• Q4: Which algorithm is used to find the shortest paths
between a given origin point and all other vertices in a
graph?
• Q5: Describe the graph coloring problem and common
variants.

45
 Graph Traversal
• A graph traversal starting at vertex v will visit all vertices w that
have a path between v and w.
- If a graph is not connected, a path traversing the graph starting at vertex v will only access a
subset of the vertices of the graph. This subset is called the connection component containing v.
- If a graph contains a cycle, then a graph traversal algorithm can iterate infinitely. To avoid that
unfortunate scenario, the algorithm must mark each vertex in one hit and never visit a vertex more
than once.

46
Depth-First Search
• From a given vertex v, the depth-first search (DFS) strategy of
the graph traversal method will proceed along the path from v
as far into the graph as possible before backing up.

47
Breadth-first search
• Breadth-first search (BFS): the graph traversal strategy visits
every vertex adjacent to v that it can before accessing any
other vertices.

48
Topological sort
• A directed graph with no cycles has a natural order. It is a
linear order, called a topological order.
• The arrangement of vertices into a topological order is called
topological sorting.

49
Spanning tree
• The spanning tree of a connected undirected
graph G is a subgraph of G containing all of G's
vertices and enough of its edges to form a tree.

51
Minimum spanning tree
• Imagine that a developing country hires you to design their
phone system so that all cities in the country can call each
other.

Central office

52
Exercises
• Draw the minimum spanning tree of

55
The shortest path
• Find the shortest path between two specific peaks in the city
map.

Small section of Williamsburg in Brooklyn, NY

56
http://www.marcinkossakowski.com/finding-shortest-path-using-dijkstras-algorithm/
 Dijkstra algorithm
• The Dijkstra algorithm determines
the shortest paths between a
given origin point and all other
vertices.
• The algorithm uses a set of
vertices The set of selected
vertices and an array weight,
where the weight [v] is the weight
of the shortest path from vertex 0
to vertex v.

59
Graph coloring
• Graph coloring is an assignment of "colors" to the
elements of a graph that satisfies given constraints.
• Some common coloring problems are:
• Vertex coloring, where adjacent vertices must not share the
same color.
• Edge coloring, where adjacent edges must not share the same
color.
• Face coloring of a planar graph, where each face or region is
colored such that no two faces sharing a boundary have the
same color.

60
Exercises
• The minimum number of colours required to colour the
following graph, such that no two adjacent vertices are
assigned the same colour, is

A) 2
B) 3
C) 4
D) 5

61
Exercises
There are 6 football teams: A, B, C, D, E, F competing in a
round-robin (one round) tournament. The following
matches have already taken place:
• Team A has played against teams B and E.
• Team B has played against teams A and F.
• Team C has played against teams D and F.
Each team can only play one match per week. Arrange
the matches into weeks so that the total number of
weeks is minimized.
62
Content
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

63
Review questions
• Q1: What is a cut in a graph?
• Q2: What is a cut-set in a graph?
• Q3: How is the size/weight of a smallest/largest slice
defined?
• Q4: What does flow represent in a graph?
• Q5: Define maximum flow in a graph.
• Q6: What is the relationship between maximum flow
and minimum slice between two vertices s and t?

64
Cut
• A cut is a way of dividing the set of vertices of a
graph into 2 subsets that do not intersect.
• A cut-set is a set of edges that each end is in a subset
after cutting.

65
Smallest slice
• A slice is smallest when its size/weight is not greater
than any other slice.

The smallest size is 2 66

The largest slice

• A slice is largest when its size/weight is not less than

any other slice.

The largest size is 4 67

Exercies
• Q1: Let G be the graph represented in figure. Is it
true that the set {ae, ef, f j, jk, cd, dh} is a cut?

68
Flow
• Flow represents the load capacity of an edge in a
graph.

The flow from s to t with a total volume of 5.

69
Maximum flow
• The maximum flow is the flow that can achieve a
maximum load transfer rate.
• The maximum value of a flow from s to t is equal to
the minimum slice between s and t.

70
Contents
• Graph concepts and properties
• Types of graphs
• Graph storage
• Some basic algorithms on graphs
‒ Graph Traversal Algorithm
‒ Topological Sorting Algorithm
‒ Spanning Tree Algorithm
‒ Shortest Path Algorithm
• Slices and Threads
• Graph Match

71
Review Questions
• Q1: Define graph isomorphism.
• Q2: What is graph edit distance?

72
Match graphs - isomorphic
• Two graphs G and H are called isomorphic if there exists a
parallelism:
𝑓: 𝑉(𝐺) → 𝑉(𝐻)

and if there is an edge between two vertices you and v in G, there will
also be an edge between the vertices f(u) and f(v) in H.

73
Exercises
• Which of the following graphs is isomorphic to:

74
Match graphs - adjustable distance

• Graph edit distance is the number of changes required

to convert one graph into the other.

75
Reference
• Mihalcea, Rada, and Dragomir Radev. Graph-
based natural language processing and
information retrieval. Cambridge university
press, 2011.
• Frank M. Carrano, Data Abstraction and Problem
Solving with C++: Walls and Mirrors, Frank M.
Carrano, 6th Edition
• Minimum Spanning Tree in Graph - Week 13. 2
Problem: Laying Telephone Wire Central office.
76