Characterization of users of

Online Social Networks

Debesh Majumdar
2nd Year Student
VGSOM, IIT Kharagpur April 29, 2011
Project Approach

• Literature review from major journals

• Secondary sources of data
• Primary data source
• Observations
• References
Why OSN?
Characteristics of OSNs

• User based
o Built and directed by users themselves
o Users populate the network with content and conversations
o Direction of content is determined by anyone who takes part in
the discussion
o Unpredictability of direction makes it dynamic and exciting
• Interactive
o Presence of network based gaming applications alongside forums
and chatrooms
o These applications allow users to connect with other users and
have fun
• Community-driven
o Users having common beliefs or interests create communities
o New connections between users are formed based on these
commonalities
Characteristics of OSNs

• Relationships
o Number of connections of users determine the user's distance
from the centre of the network
o Content published by a user proliferates exponentially as the
number of contacts increases
• Emotion over content
o OSNs provide an easy way to reach out to a large number of
friends
o Certain sense of emotional security as friends are within easy
reach
o Ability to communicate inner feelings with friends provides a
great deal of support
Classification of users

• Singletons
o Degree-zero nodes who have joined the service but have never
made a connection with another user
o Do not actively participate in the network
• Giant components
o Represent the large group of people who are connected to one
another through paths in the network
o Typically connected directly or indirectly to a large fraction of the
entire network
o Contains most of the highly active and gregarious individuals
• Middle region
o Consists of various isolated communities
o Small groups who interact with one another but not with the
network at large
o Represents a significant fraction of the entire population
Insights into evolution of OSNs

• Likelihood that two isolated communities will merge together

is low
• Almost all isolated communities are stars
o A star is a single charismatic individual linked to a varying
number of other users who have very few other
connections
• Evolution of stars characterized by two processes
o Isolated communities grow one user at a time
o Either merge into the giant component or cease to exist
when the star loses focus
• Merging stars represent the outer layer of a giant component
• A group of tightly-connected active members form the core
of a giant component
• The average distance between users in a giant component
falls over time
User Content Generation

• User posting behaviour of original content shows strong

daily and weekly patterns while non-original content posting
do not exhibit any temporal pattern
• Users can be classified into three groups according to their
posting behaviour - steadily posting, occasionally posting
and inactively posting
• 20% users contribute to 80% of the total content in the
network
• User contribution for original content characterized by
stretched exponential model with different parameters. For
example, for high-quality content the distribution has a low
stretch factor indicating that it is generated by a small core
group
OSN use and personality

• OSN use can provide a lot of data about the personality of

an individual
• Extrovert individuals have a significantly higher number of
connections than introvert individuals
• Extrovert individuals demonstrate a lower use of personal
information
• Highly neurotic individuals prefer to post more photos
• People with low and high levels of agreeableness are
inclined to upload more pictures than people with a
moderate level of agreeableness
• Individuals having high degree of openness to experience
use more features from the personal information section
• Conscientious individuals have a higher number of friends
but make less use of the picture upload feature
What do users ask their social network?

Question type Question topic

Recommendation Technology

Opinion Entertainment

Factual knowledge Home & family

Rhetorical Professional

Invitation Places
Motivation for Q&A in OSN

Questions Answers

Trust Altruism

Subjectivity of question Expertise

Belief that search engines would not

Properties of question
work

Specific audience Nature of relationship

Connect socially Connect socially

Relationship Strength

• Relationship strength is modelled on the basis of the theory

of homophily which postulates
o People tend to form ties with other people who have
similar characteristics
o Stronger the ties, higher the similarities
• Elements used to measure relationship strength
o School and university attended
o Companies worked for
o Geographical location
Revenue Models of OSNs
Lengthen the tail
Fatten the tail
Drive demand down the tail
Disclosures

• Personal information can be classified into

o Default/standard
o Sensitive
o Potentially stigmatizing
• Approximately 25% of all possible information that could
potentially be disclosed by individuals are disclosed
• Disclosure of personal information such as gender and age
is related to disclosure of other sensitive and highly personal
information
• Younger people have more personal information on their
profiles
• Users seeking a relationship tend to disclose the greatest
amount of highly sensitive information
Privacy taxonomy

• Purpose
• Specification of legitimate reasons to access a specific piece of
information
• Visibility
• Who is allowed to access provided data
• Granularity
• Degree of precision that is revealed in response to a query for a
piece of data
• Retention
• The time period during which access to data should be allowed
• Registration
• Information required to identify data provider uniquely
• Networking
• Information solicited by the social network to be released to its other
users
• Content
• Actual content with which the data provider participates in network
• Activity
• Web server logs, information from cookies, as well as other means
Legends

Purpose Visibility Granularity

RSm=Reuse Same H=House S=Specific
RS=Reuse selected F=Friends P=Partial
RA=Reuse Any FoF=Friends of friends
A=Any N=Network
A/W=All/World
Purpose

Purpose LinkedIn Twitter Orkut Facebook MySpace YouTube

Registration RA RA RSm RA RS RA

Networking A A A RAA A A

Content A A A RAA A A

Activity RS RS A RA A A
Visibility

Visibility LinkedIn Twitter Orkut Facebook MySpace YouTube

Registration H H H H H H

Networking FAW FAW FN HN FAW AW

Content FAW FAW FN HN FAW AW

Activity H H H H H H
Granularity

Visibility LinkedIn Twitter Orkut Facebook MySpace YouTube

Registration S S S S S S

Networking S,P S,P S S S S

Content S S,P S S S S

Activity P P S S S S
Facebook Applications & Privacy

• What information does Facebook share with applications?

o Practically everything ...
o Application Terms of Service
 "Facebook may...provide developers access to...your name, your profile
picture, your gender, your birthday, your hometown location...your current
location...your political view, your activities, your interests...your
relationship status, your dating interests, your relationship interests, your
summer plans, your Facebook user network affiliations, your education
history, your work history,...copies of photos in your Facebook Site photo
albums...a list of user IDs mapped to your Facebook friends."
o If you set your profile to private and one of your friends adds an
application, most of your profile information that is visible to your friend is
also available to the application developer -- even if you yourself have not
installed the application
o
Source: http://news.cnet.com/8301-13739_3-9854409-46.html#ixzz1KPfjrGfo
Information needs of Facebook apps

• 90.7% applications being given more privileges than they

need

Source : http://www.cs.virginia.edu/felt/privacy/
 Facebook's changing privacy policies
2005
"No personal information that you submit to the Facebook will be available to any user of the Web
Site who does not belong to at least one of the groups specified by you in your privacy settings."

Implications
• We are not giving your data to anyone who is not related to you in some way
 Facebook's changing privacy policies
2006
"We understand you may not want everyone in the world to have the information you share on
Facebook; that is why we give you control of your information. Our default privacy settings limit the
information displayed in your profile to your school, your specified local area, and other reasonable
community limitations that we tell you about."

Implications
• We will provide everyone will some of the details you add to your profile page.
 Facebook's changing privacy policies
2007
"Profile information you submit to Facebook will be available to users of Facebook who belong to at
least one of the networks you allow to access the information through your privacy settings (e.g.,
school, geography, friends of friends). Your name, school name, and profile picture thumbnail will be
available in search results across the Facebook network unless you alter your privacy settings."

Implications
• Apart from some of the profile information becoming public, we will allow people to search for
you and display those information
 Facebook's changing privacy policies
November 2009
"Facebook is designed to make it easy for you to share your information with anyone you want. You
decide how much information you feel comfortable sharing on Facebook and you control how it is
distributed through your privacy settings. You should review the default privacy settings and change
them if necessary to reflect your preferences. You should also consider your settings whenever you
share information. ...
Information set to “everyone” is publicly available information, may be accessed by everyone on the
Internet (including people not logged into Facebook), is subject to indexing by third party search
engines, may be associated with you outside of Facebook (such as when you visit other sites on the
internet), and may be imported and exported by us and others without privacy limitations. The
default privacy setting for certain types of information you post on Facebook is set to “everyone.”
You can review and change the default settings in your privacy settings."
Implications
• Your profile data is no longer limited to Facebook users. It can be accessed by search engines
and others "without privacy limitations"
 Facebook's changing privacy policies
December 2009
"Certain categories of information such as your name, profile photo, list of friends and pages you
are a fan of, gender, geographic region, and networks you belong to are considered publicly
available to everyone, including Facebook-enhanced applications, and therefore do not have
privacy settings. You can, however, limit the ability of others to find this information through search
using your search privacy settings."
Implications
• Some of the information you provide is now public. There is no way you can have any setting to
make them private.
 Facebook's changing privacy policies
April 2010
"When you connect with an application or website it will have access to General Information about
you. The term General Information includes your and your friends’ names, profile pictures, gender,
user IDs, connections, and any content shared using the Everyone privacy setting. ... The default
privacy setting for certain types of information you post on Facebook is set to “everyone.” ...
Because it takes two to connect, your privacy settings only control who can see the connection on
your profile page. If you are uncomfortable with the connection being publicly available, you should
consider removing (or not making) the connection."

Implications
• Any interaction you have with an application or an external website will help the application or
website to access to your and your friends' profile information.
Facebook’s Privacy Policy Statement

• Word count over the years

2010
2009

2007 5443 5830

3063
2006
2313
2005
1004
Facebook's Privacy Policy Statement
• Comparison with other OSNs
Current Facebook privacy settings
Categories of OSN applications

Category Description
Allow users to declare best friends
Friend comparison
and compare friend traits

Allow users to exchange messages

Casual communication
and write on each other's wall

Enables users to review, compare

Rating/Recommendation
and recommend items

Allows users to perform virtual

Gestures
gestures like poke and bite

Enables users to express moods,

Self expression
political opinion etc

Gifting Enables users to exchange gifts

Average Age Distribution
Age Distribution per site
Average user per site
Average user per site
Observations

• The average social network user is 37 years old

• LinkedIn has an average user age of 44
• The average Twitter user is 39 years old
• The average Facebook user is 38 years old
• Most importantly, unlike popular perception, OSNs are
dominated by middle-aged people and not by the tech-savvy
younger generation
Social Games – Look who’s playing
Why are people playing?
Disclosures on OSNs
Who is spending the most time on OSN?
Profile of virtual world users
The Hypothesis

H0 : Classification based on number of years in Facebook and

the number of hours spent each week on Facebook are
independent

H1 : Classification based on number of years in Facebook and

the number of hours spent each week on Facebook are not
independent
The Hypothesis

Number of hours Total Expected cell

spent per week frequencies per
null hypothesis
Less More
than 5 than 5
hours hours
Number More 11 42 53 15.55 37.45
of years than 1
year
Less 11 11 22 6.45 15.55
than 1
year
Total 22 53 75
The Tests

Chi X2
Square For
97.5%
6.41 5.02

Therefore we can say that number of years on FB and

number of hours spent are not independent with a
confidence level of 97.5%
References

• "Structure & evolution of Online Social Networks"

o Ravi Kumar, Jasmine Novak, Andrew Tomkins
• "Social network use and personality"
o Yair Amichai-Hamburger, Gideon Vinitzky
• "Modelling relationship strength in Online Social Networks"
o Rongjing Xiang, Jennifer Neville, Monica Rogati
• "All about me: Disclosure in Online Social Networking profile :
The case of Facebook"
o Amanda Nosko, Eileen Wood, Seija Molema
• "Analyzing patterns of user content generation in Online
Social Networks"
o Lei Guo, Enhua Tan, Songqing Chen, Xiaodong Zhang, Yihong Zhao
• The long tail of social networking sites
• Albrecht Enders, Harald Hungenberg, Hans-Peter Denker
• What do people ask their social networks and why?
o Meredith Ringel Morris, Jaime Teevan, Katrina Panovich
Thank You!