1.
© Copyright 2013 INSIGHT Centre for Data Analytics. All rights
reserved.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
“I Like” – Analysing Interactions within Social
Networks to Assert the Trustworthiness
of Users, Sources and Content
Owen Sacco & John G. Breslin
owen.sacco@deri.org & john.breslin@nuigalway.ie
ASE/IEEE SocialCom 2013
Washington, DC, USA
Tuesday 10th
September 2013

2.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Introduction
 Current Social Networks:
 Provide generic privacy settings for sharing information
 Do not take user’s trust into account

3.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Introduction
In reality, we only share parts of our
information with whomever we trust

4.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Social Factors
 Trust judgments are influenced by Social Factors:
 Past interactions with a person
 Opinions of a person’s actions
 Other people’s opinions
 Rumours
 Psychological factors impacted over time
 Life events
 and so forth
 These can be hard to compute since the
information required is limited and unavailable in
Social Networks

5.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Research Questions
1. What is the user’s perception of trust?
2. Which information extracted from Social
Networks is useful for computing trust?
3. For what and for whom can trust be
computed from the information in
Social Networks?

6.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey
 To analyse how trust can be inferred from Social
Networks
 We focus on whether trust can be asserted from
user interactions within the Social Networks
 User interactions with:
 Other users
 Content shared within the Social Networks

7.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey
 User interactions:
 Sharing of content from external sources
 Re-sharing or retweeting content
 “Like”, or “+1” or “favourite”
 Comments or replies
 Tags or mentions

8.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey
 178 participated in the survey: 65% male and 35% female
 Age:
 Social Network Accounts:
Age Category Participants
18 - 20 3%
21 - 29 45%
30 - 39 32%
40 - 49 13%
50 - 59 6%
60+ 1%
Social Networks Participants
Facebook 88%
Google+ 69%
Twitter 82%
LinkedIn 85%
None of the Above 1%

9.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey
 Occupations of participants
Occupation Categories Participants
Computer and Mathematics 59%
Education, Training and Library 26%
Business and Financial 13%
Management 8%
Architecture and Engineering 6%
Arts, Design, Entertainment, Sports and Media 5%
Life, Physical and Social Sciences 3%
Office and Administrative Support 2%
Healthcare Support 1%
Community & Social Service 1%
Sales 1%
Unemployed 1%

10.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey
 Two parts:
1. Usage Patterns
2. User’s Trust Perception in Social Networks
 Usage Patterns: how often users use each social
user interaction and on which Social Network
 User’s Trust Perception: analyses what users trust
when they use these social user interactions

11.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
 How often do you share content from external
sources within Facebook, Google+, Twitter and
LinkedIn?
User Survey:
Usage Patterns

12.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Usage Patterns
 How often do you re-share or retweet what other
users share within Facebook, Google+, Twitter and
LinkedIn?

13.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Usage Patterns
 How often do you use the like, +1 and favourite
buttons on Facebook, Google+, Twitter and
LinkedIn?

14.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Usage Patterns
 For what do you use the like, +1 and favourite
buttons on Facebook, Google+, Twitter and
LinkedIn?

15.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Usage Patterns
 How often do you comment or reply on Facebook,
Google+, Twitter and LinkedIn?

16.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Usage Patterns
 How often do you tag or mention other users on
Facebook, Google+, Twitter and LinkedIn?

17.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What is your perception of the meaning of the word
“Trust”?

18.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you share external content
into Facebook, Google+, Twitter and LinkedIn?

19.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you re-share or retweet
content within Facebook, Google+, Twitter and
LinkedIn?

20.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you like, +1, or favourite
within Facebook, Google+, Twitter and LinkedIn?

21.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you comment or reply to
posts within Facebook, Google+, Twitter and
LinkedIn?

22.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you tag or mention other
users within Facebook, Google+, Twitter and
LinkedIn?

23.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
User’s Trust Perception
 What do you trust when you are tagged or
mentioned within Facebook, Google+, Twitter and
LinkedIn?

24.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Summary
 Overall Participants’ Activity of Social User
Interactions

25.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
User Survey:
Summary
 Overall Participants’ Perception of Trust

26.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Asserting Trust:
Trusting the Source
 Trust for the source can be asserted from:
 The share button
 The re-share or retweet buttons
 Trusting the source:
» denotes the user’s subjective trust value for a particularτ
source
» w denotes the trust a third party user has in the user’s social
graph
» s denotes the number of shares and re-shares related to the
source

27.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Asserting Trust:
Trusting the Content
 Trust for content can be asserted from:
 The share button
 The re-share or retweet buttons
 The like, +1 and favourite buttons
 Trusting the content:
» denotes the user’s subjective trust value for a particularτ
content
» w denotes the trust value a third party user has in the user’s
social graph
» c denotes the number of shares, re-shares, likes, +1s and
favourites related to the content

28.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Asserting Trust:
Trusting the User
 Trust for the user (i.e. requester) can be asserted
from:
 The like, +1 and favourite buttons
 The comments or replies to posts
 The tags of the requester tagged by the information owner
 The tags of the information owner tagged by the requester
 Trusting the user (i.e. requester):
» denotes the user’s subjective trust value for a particularτ
requester
» r denotes the number of likes, +1s, favourites, comments,
replies and tags related to the user and the requester

29.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Conclusion & Future Work
 We focused on:
 Analysing the user’s perception of trust and
 How trust can be inferred from Social Networks
 User survey that analysed the usage patterns and
the user’s trust perception of:
 The share button
 The re-share or retweet buttons
 The like, +1 or favourite buttons
 The comment or reply buttons
 The tag or mention buttons/options

30.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Conclusion & Future Work
 The results have revealed that users are concerned
with asserting trust for:
 The source that created the content
 The content
 The user requesting personal information
 Future work:
 Implementing the trust assertions in our Privacy
Preference Framework (see previous publications)
– To enforce privacy preferences based on these trust
assertions

31.
INSIGHT Centre for Data Analytics www.insight-centre.org
Semantic Web & Linked Data
Research Programme
Thanks!
@owensacco
owen.sacco@deri.org
@johnbreslin
john.breslin@nuigalway.ie