School of Engineering
Emotional Analysis in Social Networks: A Study in
Application of Techniques & Methods for Capturing it
Albert Savill
Student Number: 3505901
BSc (Hons) Computer Science
London South Bank University
May 2019
,Declaration
Abstract
This project is analysing the emotions in social networks and how users in social networks
can have an emotional influence on others. In this project, I will be analysing emotions of
social network users under certain events, where from these certain events there is guaranteed
to have a mix of emotions. From the measurements of emotions found, I can also find out
how some users can influence other user’s emotions based on their following.
The importance of this topic is to deepen the understanding of what keywords in social
networks can determine the emotions of the users. There are many reasons why learning the
typical emotions of users in social networks are important. For example, it can be used for
businesses to understand the emotions of their customers better regarding trends and
products. Learning emotions can also help us detect which users are in need of mental help
based on the words they express in their statuses or tweets.
i|Page
,The following method of this project would be SentiStrength, a method inspired by other
research projects (Thelwall, Buckley, & Paltoglou, 2012) where I will use SentiStrength
detection in the social web, then go on to see the influence methods used by people on social
media.
Acknowledgements
I would like to thank my supervisor Christos Chrysoulas for allowing me to undertake this
project. I would also like to thank him for the support and guiding me to the correct place to
look for all of those research journals that was the gateway to my research. I would also like
to thank him for being a great lecturer from previous modules over the years.
I would also like to thank my second supervisor Mike Child for getting involved in this
project by meeting me for the mid-project report. I would also like to thank him for being a
great lecturer from previous modules over the years.
Also, I would like to thank Professor Mike Thelwall from the University of Wolverhampton
for allowing me to use his sentiment strength java manual guide which helped me code the
sentiment strength program.
Also, a thank you to every academic who wrote journals or articles (cited in references).
Without them, I would have had no idea about this area and certainly failed my final year
project.
Finally, I would like to thank my classmates but more importantly, my friends who
undertaken the final year dissertation at the same time I did. It is always nice to crack jokes
together underneath relatable/stressful times during our studies over the years.
Table of Contents
ii | P a g e
, Declaration.............................................................................................................................................i
Abstract..................................................................................................................................................i
Acknowledgements...............................................................................................................................ii
Table of Figures.....................................................................................................................................0
Chapter 1 Discussion of Topic................................................................................................................1
1.1 Topic............................................................................................................................................1
1.2 Importance of topic.....................................................................................................................2
1.3 Followed method.........................................................................................................................3
Chapter 2 Extended Literature Review..................................................................................................5
2.1 Historical facts.............................................................................................................................5
2.2 Used Terminology regarding Emotional Analysis in Social Networks...........................................7
2.2.1 Basic terminology..................................................................................................................7
2.2.2 Technical terminology...........................................................................................................8
2.3 Breakthrough in the Social Network Emotional Analysis.............................................................9
2.4 Future........................................................................................................................................10
Chapter 3 Proposed Method...............................................................................................................11
3.1 Introduction...............................................................................................................................11
3.2 Calculating the emotional importance of tweets.......................................................................12
3.3 How to measure the influence of the users...............................................................................13
Chapter 4 Implementation..................................................................................................................18
4.1 Introduction...............................................................................................................................18
4.2 Used Tools and Framework.......................................................................................................18
4.3 Architecture...............................................................................................................................20
4.4 Raw results presentation and initial discussion.........................................................................21
Chapter 5 Analysing results.................................................................................................................23
5.1 Introduction...............................................................................................................................23
5.2 Analysing and comparing the results.........................................................................................24
5.2.1 Normal sentiment tweet results.........................................................................................24
5.2.2 Low scoring sentiment tweets............................................................................................26
5.2.3 Equal Sentiment Tweet.......................................................................................................26
5.2.4 Tweets With Sentiment Boosters........................................................................................27
5.2.5 Outlier Tweets Through Wrong Sentiment Score...............................................................28
5.3 Checking the influence of the users in the past.........................................................................29
5.3.1 Celebrity influence..............................................................................................................30
5.3.2 Political Journalist...............................................................................................................32
5.3.3 Actual Politicians.................................................................................................................33
iii | P a g e
Emotional Analysis in Social Networks: A Study in
Application of Techniques & Methods for Capturing it
Albert Savill
Student Number: 3505901
BSc (Hons) Computer Science
London South Bank University
May 2019
,Declaration
Abstract
This project is analysing the emotions in social networks and how users in social networks
can have an emotional influence on others. In this project, I will be analysing emotions of
social network users under certain events, where from these certain events there is guaranteed
to have a mix of emotions. From the measurements of emotions found, I can also find out
how some users can influence other user’s emotions based on their following.
The importance of this topic is to deepen the understanding of what keywords in social
networks can determine the emotions of the users. There are many reasons why learning the
typical emotions of users in social networks are important. For example, it can be used for
businesses to understand the emotions of their customers better regarding trends and
products. Learning emotions can also help us detect which users are in need of mental help
based on the words they express in their statuses or tweets.
i|Page
,The following method of this project would be SentiStrength, a method inspired by other
research projects (Thelwall, Buckley, & Paltoglou, 2012) where I will use SentiStrength
detection in the social web, then go on to see the influence methods used by people on social
media.
Acknowledgements
I would like to thank my supervisor Christos Chrysoulas for allowing me to undertake this
project. I would also like to thank him for the support and guiding me to the correct place to
look for all of those research journals that was the gateway to my research. I would also like
to thank him for being a great lecturer from previous modules over the years.
I would also like to thank my second supervisor Mike Child for getting involved in this
project by meeting me for the mid-project report. I would also like to thank him for being a
great lecturer from previous modules over the years.
Also, I would like to thank Professor Mike Thelwall from the University of Wolverhampton
for allowing me to use his sentiment strength java manual guide which helped me code the
sentiment strength program.
Also, a thank you to every academic who wrote journals or articles (cited in references).
Without them, I would have had no idea about this area and certainly failed my final year
project.
Finally, I would like to thank my classmates but more importantly, my friends who
undertaken the final year dissertation at the same time I did. It is always nice to crack jokes
together underneath relatable/stressful times during our studies over the years.
Table of Contents
ii | P a g e
, Declaration.............................................................................................................................................i
Abstract..................................................................................................................................................i
Acknowledgements...............................................................................................................................ii
Table of Figures.....................................................................................................................................0
Chapter 1 Discussion of Topic................................................................................................................1
1.1 Topic............................................................................................................................................1
1.2 Importance of topic.....................................................................................................................2
1.3 Followed method.........................................................................................................................3
Chapter 2 Extended Literature Review..................................................................................................5
2.1 Historical facts.............................................................................................................................5
2.2 Used Terminology regarding Emotional Analysis in Social Networks...........................................7
2.2.1 Basic terminology..................................................................................................................7
2.2.2 Technical terminology...........................................................................................................8
2.3 Breakthrough in the Social Network Emotional Analysis.............................................................9
2.4 Future........................................................................................................................................10
Chapter 3 Proposed Method...............................................................................................................11
3.1 Introduction...............................................................................................................................11
3.2 Calculating the emotional importance of tweets.......................................................................12
3.3 How to measure the influence of the users...............................................................................13
Chapter 4 Implementation..................................................................................................................18
4.1 Introduction...............................................................................................................................18
4.2 Used Tools and Framework.......................................................................................................18
4.3 Architecture...............................................................................................................................20
4.4 Raw results presentation and initial discussion.........................................................................21
Chapter 5 Analysing results.................................................................................................................23
5.1 Introduction...............................................................................................................................23
5.2 Analysing and comparing the results.........................................................................................24
5.2.1 Normal sentiment tweet results.........................................................................................24
5.2.2 Low scoring sentiment tweets............................................................................................26
5.2.3 Equal Sentiment Tweet.......................................................................................................26
5.2.4 Tweets With Sentiment Boosters........................................................................................27
5.2.5 Outlier Tweets Through Wrong Sentiment Score...............................................................28
5.3 Checking the influence of the users in the past.........................................................................29
5.3.1 Celebrity influence..............................................................................................................30
5.3.2 Political Journalist...............................................................................................................32
5.3.3 Actual Politicians.................................................................................................................33
iii | P a g e
