The purpose of this study is to apply social network analysis techniques to construct a social network out of an email dataset, as well as analyze and visualize the network properties and utilize sentiment analysis as an additional source of information to study employees relationships in a company. This study is primarily based on the Enron email dataset and covers the methodology followed to transform the data into a suitable format to detect patterns and get useful information. The social graph is based on the From and To fields in the data, plus the distribution of emails sent by the entities. The resulting social graph contains around 371 nodes and 67 thousand edges, with a ratio of 84% neutral messages, 11% positive and approximately 5% negative. It was concluded that when social network analysis is used in conjunction with emotion detection, it is possible to see the positive or negative areas where the company must work in order to promote a healthy organizational culture and uncover possible organizational issues in a timely manner.

本研究的目的是應用社交網絡分析技術來構建社會網絡出來的電子郵件數據集，以及分析和可視化的網絡屬性和利用情緒分析的信息的附加源，研究在公司員工的關係。這項研究主要是基於安然郵件數據集，涵蓋了遵循的方法將數據轉換成合適的格式來檢測模式，並獲取有用的信息。社交圖是基於來自與要在數據字段，加上由實體發送的電子郵件的分發。由此產生的社會圖包含圍繞371的節點和67000的邊緣，用84％中性消息，11％正和大約5％的負的比率。得出的結論是，當社會網絡分析與情緒檢測結合使用時，就可以看到該公司必須以促進健康的組織文化和及時發現可能的組織問題的工作積極或消極的方面。

Contents
Abstract iii
Acknowledgements iv
List of Figures vi
List of Tables vii
1 Introduction 1
2 Related Work 2
3 Methodology 5
3.1 The Enron Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2 Data transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.3 Extracting the email addresses . . . . . . . . . . . . . . . . . . . . . . . . 6
3.4 Preprocessing the email addresses . . . . . . . . . . . . . . . . . . . . . . 6
3.5 Preprocessing the messages . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.6 Feature selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.7 Sentiment analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.8 Social Network Analysis on the Enron emails . . . . . . . . . . . . . . . . 13
4 Results and Discussion . . . . . . . . . . . . . . . . . . . . .14
4.1 The documents in the Enron Corpus . . . . . . . . . . . . . . . . . . . . . 14
4.2 Enron Emails Sentiment Classification . . . . . . . . . . . . . . . . . . . . 15
4.3 Social network Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3.1 Degree Centrality . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3.2 Betweenness Centrality . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3.3 Closeness Centrality . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.4 Sentiment diffusion in social network . . . . . . . . . . . . . . . . . . . . 23
4.4.1 Positive sentiment subgraph . . . . . . . . . . . . . . . . . . . . . 23
4.4.2 Negative sentiment subgraph . . . . . . . . . . . . . . . . . . . . . 24
4.4.3 Neutral sentiment subgraph . . . . . . . . . . . . . . . . . . . . . 25
5 Conclusions . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . .27
References . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . .29

References
[1] Mohammed J. Zaki and Wagner Meira Jr. Data Mining and Analysis: Fundamental
Concepts and Algorithms. New York, NY, USA: Cambridge University Press, 2014.
ISBN: 0521766338, 9780521766333.
[2] Bethany McLean and Peter Elkind. The smartest guys in the room: the amazing rise
and scandalous fall of Enron. New York, NY, USA: Penguin UK Business and Economics,
2013.
[3] Jitesh Shetty and Jafar Adibi. “Discovering Important Nodes Through Graph Entropy
the Case of Enron Email Database”. Proceedings of the 3rd International Workshop
on Link Discovery. LinkKDD ’05. Chicago, Illinois: ACM, 2005, pp. 74–81. ISBN:
1-59593-215-1. DOI: 10.1145/1134271.1134282. URL: http://doi.acm.org/
10.1145/1134271.1134282.
[4] G. Wilson and W. Banzhaf. “Discovery of email communication networks from the
Enron corpus with a genetic algorithm using social network analysis”. 2009 IEEE
Congress on Evolutionary Computation. 2009, pp. 3256–3263. DOI: 10.1109/CEC.
2009.4983357.
[5] Jana Diesner, Terrill L. Frantz, and Kathleen M. Carley. “Communication Networks
from the Enron Email Corpus “It’s Always About the People. Enron is no Different””.
Computational & Mathematical Organization Theory 11.3 (2005), pp. 201–
228. ISSN: 1572-9346. DOI: 10.1007/s10588-005-5377-0. URL: http://dx.
doi.org/10.1007/s10588-005-5377-0.
[6] B. Collingsworth, R. Menezes, and P. Martins. “Assessing organizational stability
via network analysis”. 2009 IEEE Symposium on Computational Intelligence for
Financial Engineering. 2009, pp. 43–50. DOI: 10.1109/CIFER.2009.4937501.
[7] Z. Hailong, G. Wenyan, and J. Bo. “Machine Learning and Lexicon Based Methods
for Sentiment Classification: A Survey”. Web Information System and Application
Conference (WISA), 2014 11th. 2014, pp. 262–265. DOI: 10.1109/WISA.2014.55.
[8] E. Cambria et al. “New Avenues in Opinion Mining and Sentiment Analysis”. IEEE
Intelligent Systems 28.2 (2013), pp. 15–21. ISSN: 1541-1672. DOI: 10.1109/MIS.
2013.30.
[9] X. Yu et al. “Mining Online Reviews for Predicting Sales Performance: A Case Study
in the Movie Domain”. IEEE Transactions on Knowledge and Data Engineering 24.4
(2012), pp. 720–734. ISSN: 1041-4347. DOI: 10.1109/TKDE.2010.269.
[10] Matthew A. Russell. Mining the Social Web: Analyzing Data from Facebook, Twitter,
LinkedIn, and Other Social Media Sites. 1st. O’Reilly Media, Inc., 2011. ISBN:
9781449388348.
[11] Alper Kursat Uysal and Serkan Gunal. “The impact of preprocessing on text classification”.
Information Processing and Management 50.1 (2014), pp. 104 –112. ISSN:
0306-4573. DOI: http://dx.doi.org/10.1016/j.ipm.2013.08.006. URL:
http://www.sciencedirect.com/science/article/pii/S0306457313000964.
[12] Ingo Feinerer, Kurt Hornik, and David Meyer. “Text Mining Infrastructure in R”.
Journal of Statistical Software 25.5 (2008), pp. 1–54. URL: http://www.jstatsoft.
org/v25/i05/.
[13] José M. Carmona-Cejudo et al. “A Comparative Study on Feature Selection and
Adaptive Strategies for Email Foldering Using the ABC-DynF Framework”. Know.-
Based Syst. 46 (July 2013), pp. 81–94. ISSN: 0950-7051. DOI: 10.1016/j.knosys.
2013.03.006. URL: http://dx.doi.org/10.1016/j.knosys.2013.03.006.
[14] Tim O’Keefe and Irena Koprinska. “Feature selection and weighting methods in sentiment
analysis”. Proceedings of the 14th Australasian document computing symposium,
Sydney. Citeseer. 2009, pp. 67–74.
[15] K. Schouten and F. Frasincar. “Survey on Aspect-Level Sentiment Analysis”. IEEE
Transactions on Knowledge and Data Engineering 28.3 (2016), pp. 813–830. ISSN:
1041-4347. DOI: 10.1109/TKDE.2015.2485209.
[16] Bo Pang and Lillian Lee. “Opinion Mining and Sentiment Analysis”. Found. Trends
Inf. Retr. 2.1-2 (Jan. 2008), pp. 1–135. ISSN: 1554-0669. DOI: 10.1561/1500000011.
URL: http://dx.doi.org/10.1561/1500000011.
[17] Amber E. Boydstun Emiliano Grossman Timothy P. Jurka Loren Collingwood and
Wouter van Atteveldt. “RTextTools: Automatic Text Classification via Supervised
Learning” (2012). URL: http://CRAN.R-project.org/package=RTextTools.
[18] Charu C Aggarwal. “An introduction to social network data analytics”. Social network
data analytics. Springer, 2011, pp. 1–15.