Unit III

Social Network
Analysis
BSc Data Science
Course
By Pushpa Mahapatro
 NETWORK

At a very basic level, a network is a group of nodes

Introduction to that are connected with links

Networks: Nodes (also known as vertices) can represent

anything, including individuals, organizations,
countries, computers, websites, or any other entities.

Links (also known as ties, edges, or arcs) represent the

relationship among the nodes in a network.
SOCIAL NETWORKS
•A social network is a group
of nodes and links formed
by social entities where
nodes can represent social
entities such as people
and organizations.
SOCIAL NETWORK SITE:

A social network site is designed to

facilitate the creation and
maintenance of social relations.

Facebook, Google+, and LinkedIn

are examples of social network
sites.
SOCIAL NETWORKING:
• The act of forming, expanding, and
maintaining social relations is called
social networking.

SOCIAL NETWORK ANALYSIS:

• Social network analysis is the
science of studying and
understanding social networks and
social networking.
Network
structure
• The following are some
everyday types of social
media networks that we
come across and that can
be subject to network
analytics.
FRIENDSHIP NETWORKS

friendship networks, such as Facebook, Google+,

and Cyword.

Friendship networks let people maintain social ties

and share content with people

Nodes in these networks are people, and links are

social relationships (e.g., friendship, family, and
activities).
FOLLOW-FOLLOWING NETWORKS

In the follow-following network, users follow (or

keep track of) other users of interested.

Nodes in these networks are, for example, people,

brands, and organizations, and links represents
follow-

following relations (e.g., who is following whom).

Below are two common Twitter terminologies.
FAN NETWORK:
• A fan network is formed by social media fans or supporters of
someone or something, such as a product, service, person,
brand, business, or other entity.
GROUP NETWORK:
• Group networks are formed by people who share common
interests and agendas. Most social media platforms allow the
creation of groups where member can post, comment, and
manage in-group activities.
PROFESSIONAL NETWORKS
• LinkedIn is a good example of professional networks where
people manage their professional identify by creating a profile
that lists their achievements, education, work history, and
interests.
CONTENT NETWORKS:
• Content networks are formed by the content posted by social
media users. A network among YouTube videos is an example of
a content network.
COAUTHORSHIP NETWORKS
• Coauthorship networks are two or more people working together to
collaborate on a project.
COCOMMENTER NETWORKS
• Cocommenter networks are formed when two or more people
comment on social media content (e.g., a Facebook status update,
blog post, or YouTube video).
COOCCURRENCE NETWORK
• Cooccurrence networks are formed when two more entities (e.g.,
keywords, people, ideas, and brands) cooccur over social media
outlets.
GEO COEXISTENCE NETWORK
• Geo coexistence networks are formed when two more entities
(e.g., people, devices, and addresses) coexist in a geographic
location. HYPERLINK NETWORKS
 From a technical point of view, the above-mentioned
networks can be classified in a variety of ways, including
Types of Networks
1) based on existence,

2) based on direction of links,

3) based on mode, and

4) based on weights.
 Based on the way the networks exist online or are

BASED ON EXISTENCE:
constructed, they can be classified as 1) implicit
networks or 2) explicit networks.

Implicit Networks

• Implicit networks do not exit by default (or are hidden) and

need to be intentionally constructed with the help of dedicated
tools and techniques.

Explicit Networks

• Explicit social media networks exist by default; in other words,

they are explicitly designed for social media users to be part of.
 Based on the directions of links among
the nodes, the networks can be

BASED ON DIRECTION
classified as 1) directed networks, and
2) undirected networks.

Directed Networks

Undirected Network
 Based on the composition of nodes,
networks can be classified as 1) one-
BASED ON MODE mode network, 2) two-mode
networks, and 3) multimode networks.
• One-Mode Networks: A one-mode network is
formed among a single set of nodes of the same
nature (Figure 6).
• Two-Mode Networks: Two-mode networks (also
known as bipartite networks) are networks with
two sets of nodes of different classes.
• Multimode Network: A multimode network is
also possible where multiple heterogeneous
nodes are connected together.
BASED ON WEIGHTS

• Networks can also be classified based the weight

assigned to the links among the nodes. Mainly
there are two types of weighted networks:

1) weighted networks, and 2) unweighted

networks.
• Weighted Networks: In weighted networks, the
links among nodes bear certain weights to indicate
the strength of association among the nodes.
• UNWEIGHTED NETWORKS: In unweighted
networks, links among nodes does not bear
weights.
Figure 6. Types of social media networks
Egocentric networks
• Ego-centric networks (or shortened to “ego”
networks) are a particular type of network which
specifically maps the connections of and from the
perspective of a single person (an “ego”).
While having data on only one person’s ego does not
tell scholars much, when we have tens, hundreds, or
thousands of ego-networks we are able to analyze how
different types of people have different ego-networks.

These are empirical questions which, when scholars

creatively compare ego-centric networks, we are able
to potentially answer.

When discussing an ego network, size describes the

number of alters that the ego has identified.
• Just as with a complete graph, we might be interested in the certain
properties of the ego-network, such as how dense it is. By definition,
everyone knows the ego, but to what extent does someone’s friends
know each other?
Thus, as demonstrated by the resulting
clustering coefficient, we can
understand the 40% of all possible ties
amongst ego’s friends exist.

With multiple egos, we might thus be

able to compare their personal
networks to build theories about how
the social world operates.
 Social network metrics
measure the outcome of your
Network analysis

online actions through data.

It helps you to understand

metrics

user behavior, market trends,

and other variables.
Now that you understand what metrics are, let’s
take a look at the important ones:
1. Engagement Rate
• This data reveals how many people were
involved with your post through likes,
comments, and post sharing.
• To calculate the engagement rate, you need to
see the number of interactions (likes,
comments, shares) and divide by the number of
impressions (reach).
• Example: A Facebook post reached 4,500 people
and got 665 likes, 254 comments, and 54 shares.
Did this post get a reasonable engagement rate?
• Let’s go to the calculations:
• (665+254+54)/4500 = 0,2162 or (21,62%
engagement rate)
2. Organic Reach
• Decreasing organic reach is a big challenge
for social media analysts.
• Organic reach counts how many people
were naturally hit by your posts without
paying you a dime.
• Each social network has its algorithms and
promotion criteria, so you need to take that
into account when producing and delivering
content, primarily as social networks often
don’t read their posts organically to their
entire base.
3. Bounce Rate
• Bounce Rate can be the percentage of users
who have logged in to a page (in the case of a
website)
• This metric is important because it shows how
many people are rejecting and are bothered by
your social media posts or your website’s
landing page.
4. Click-through rate
• Click-Through Rate (CTR) is also a useful metric for
observing how well your posts are performing,
evaluating if who views your content clicks, and
interact with it.
• Let’s go to the calculations:
• 24/250=0,096 or (9,6%)
• A high CTR indicates that users liked your
content and found it useful, while a low CTR
means that you need to improve the
performance of the account you manage by
devising new action plans.
Strong and Weak Ties

Strong and weak ties are both relevant and important in your social
networking interactions.

They perform different functions in relationships but they can extend

your network far beyond your normal reach.

Think how you use LinkedIn for example. Are all of your relationships
“strong ties?” Do you count all of your connections as good friends?
Are your Facebook relationships all
strong ties, or do you stick to Dunbars
number theory? You’ll probably find
several sets of weak ties in your social
networks once you start to look.

Mark Granovetter refers to your

strong ties as your friends and your
weak ties as your acquaintances in his
paper “Notes on the strength of weak
ties”
A strong tie is someone who you
know well. You’ve probably got their
number on your phone.

There is good 2 way conversation, and

even if you don’t know everything
about them

Think about a group of geeks talking

about technology.
 A weak tie is a more
Think about a couple of
tenuous relationship. Once
your friends who
a year, you may send them
understand technology but
a Christmas message
you wouldn’t class them as
promising to be in touch
geeks.
more often.
Clustering and Grouping

Network properties provide insight into the overall structure and

health of a network.

Important network-level properties include clustering coefficient,

density, diameter, average degree, and components.
• Clustering Coefficient: The clustering coefficient of a network is the degree to
which nodes in a network tend to cluster or group together.
• Density: The density of a network deals with a number of links in a network.
Components: Components of a network are the isolated sub-networks
that connect within, but are disconnected between, sub-networks

Diameter: The diameter of a network is the largest of all the calculated

shortest path between any pair of nodes in a network

Average Degree: The average degree centrality measures the average

number of links among nodes in a network.
Thank you…