近年來社群媒體的興起，訊息經過社群網路快速傳播，使用者各種意見形成公眾輿論。有心人士企圖利用大量的假帳號，操作輿論影響多數人的想法，來達到特定的目的。輿論帶風向者往往透過寫手發文後，由真人或機器人程式，操作大量假帳號，在發文後的短時間內大量的留言，以達到帶風向、製造輿論的目的。
本論文根據使用者在社群媒體上留言的共謀行為，研究由已知的異常帳號來探索出未知的同夥異常帳號。我們運用深度學習技術以計算共謀行為的相似度。本論文以國內最大的BBS站PTT為例，實驗PTT 2018年8月至2020年2月八卦版及政黑板的資料。實驗結果顯示本論文的方法可有效地由異常帳號探索出具有協作行為的未知異常帳號。

As social media service is more and more popular, information is shared and spread quickly over the social network. Some try to manipulate the public opinion by means of malicious accounts. It has been reported that one way of public opinion manipulation can be achieved by delivering the stories, and operating large amounts of malicious accounts to promote the stories few minutes after the delivery of story in a short period of time.
According to the observation of collusive behaviors of comment operations between malicious accounts over social media, this thesis investigates the exploration by examples approach to explore unknown accomplices by the known malicious accounts. Deep learning technique is leveraged to discover the similarity of collusive behaviors. The experiments were performed based on data collected from PTT Gossiping and HatePolitics board from August 2018 to February 2020. The experimental results show that the proposed mechanism can effectively discover collusive behaviors of malicious accounts.

[1] Kayode Sakariyah Adewole, Nor Badrul Anuar, Amirrudin Kamsin, Kasturi Dewi Varathan, Syed Abdul Razak. Malicious Accounts: Dark of the Social Networks. Journal of Network and Computer Applications, Vol. 79, 2017,.
[2] Muhammad Al-Qurishi, Mabrook Al-Rakhami, Atif Alamri, Majed Alrubaian, Sk Md Mizanur Rahman, and M. Shamim Hossain. Sybil Defense Techniques in Online Social Networks- A Survey. IEEE Access, Vol. 5, 2017.
[3] Qiang Cao, Michael Sirivianos, Xiaowei Yang, and Tiago Pregueiro. Aiding the Detection of Fake Accounts in Large Scale Social Online Services. USENIX/ACM Symposium on Networked Systems Design and Implementation , 2012.
[4] Manuel Egele, Gianluca Stringhini, Christopher Kruegel, and Giovanni Vigna. Towards Detecting Compromised Accounts on Social Networks. IEEE Transactions on Dependable and Secure Computing, Vol. 14, Issue 4, 2017.
[5] Ahmed Elazab, Mahmood A. Mahmood, Hesham A. HefnyHesham, and A. Hefny. Fake Accounts Detection in Twitter Based on Minimum Weighted Feature set. International Scholarly and Scientific Research and Innovation, Vol. 10, No. 1, 2016.
[6] Rodrigo Augusto Igawa, Sylvio Barbon Jr, Kátia Cristina Silva Paulo, Guilherme Sakaji Kido, Rodrigo Capobianco Guido, Mario Lemes Proença Júnior, and Ivan Nunes da Silva. Account Classification in Online Social Networks with LBCA and Wavelets. Information Sciences, Vol. 332, 2016.
[7] Sangho Lee, and Jong Kim. Early Filtering of Ephemeral Malicious Accounts on Twitter. Computer Communications, Vol. 54, 2014.
[8] Jure Leskovec, Anand Rajaraman, and Jeffrey David Ullman. Finding Similar Items. In Mining of Massive Datasets, Cambridge University Press, 2020.
[9] Tomas Mikolov, Kai Chen, Greg S. Corrado, and Jeffrey Dean. Efficient Estimation of Word Representations in Vector Space. International Conference on Learning Representations, 2013.
[10] David L. Olson, and Dursun Delen. Performance Evaluation for Predictive Modeling. In: Advanced Data Mining Techniques, Springer, 2008.
[11] Kai Shu, Amy Sliva, Suhang Wang, Jiliang Tang, and Huan Liu. Fake News Detection on Social Media: A Data Mining Perspective. ACM SIGKDD Explorations Newsletter, Vol. 109, No. 1, 2017.
[12] Monika Singh, Divya Bansal, and Sanjeev Sofat. Detecting Malicious Users in Twitter using Classifiers. 7th International Conference on Security of Information and Networks, , 2014,.
[13] Bimal Viswanath, Muhammad Ahmad Bashir, Mark Edward, Saikat Guh, Krishna Phani Gummadi, Balachander Krishnamurthy, and Alan Mislove. Towards Detecting Anomalous User Behavior in Online Social Networks. USENIX Conference on Security Symposium, August, 23rd, 2014.
[14] Soroush Vosoughi, Deb Roy, and Sinan Aral. The Spread of True and False News Online. Science, Vol. 395, Issue 6380, 2018.
[15] Gang Wang, Manish Mohanlal, Christo Wilson, Xiao Wang, Miriam Metzger, Haitao Zheng, and Ben Y. Zhao. Social Turing Tests: Crowdsourcing Sybil Detection. The Network and Distributed System Security Symposium, The Internet Society, 2013.
[16] Ming-Hung Wang, Nhut-Lam Nguyen, Shih-chan Dai, Po-Wen Chi, and Chyi-Ren Dow. Understanding Potential Cyber-Armies in Elections: A Study of Taiwan. Sustainability, Vol. 12, No. 6, 2020.
[17] Ming-Hung Wang, Nhut-Lam Nguyen, and Chyi-Ren Dow. Detecting Potential Cyber Armies of Election Campaigns Based on Behavioral Analysis. In Complex Networks and Their Applications VII, 2018.
[18] Yahan Wang, Chunhua Wu, Kangfeng Zheng, and Xiujuan Wang. Social Bot Detection Using Tweets Similarity. Security and Privacy in Communication Networks, 2018.
[19] Haifeng Yu, Phillip B. Gibbons, Michael Kaminsky, and Feng Xiao. SybilLimit: A Near-Optimal Social Network Defense Against Sybil Attacks. IEEE/ACM Transactions on Networking, Vol. 18, No. 3, 2010.
[20] Haifeng Yu, Michael Kaminsky, Phillip B. Gibbons, and Abraham D. Flaxman. SybilGuard: Defending against Sybil Attacks via Social Networks. IEEE/ACM Transactions on Networking, Vol. 16, 2008.
[21] Xinyi Zhou, Reza Zafarani, Kai Shu, and Huan Liu. Fake News: Fundamental Theories, Detection Strategies and Challenges. 12th ACM International Conference on Web Search and Data Mining, 2019.
[22] 孔德廉，誰帶風向：被金錢操弄的公共輿論戰爭，報導者，2018/09/26。
[23] 孔德廉，網紅、假帳號、素人暗樁──值得信賴的口碑行銷？，報導者，2018/09/26。
[24] 林倖妃，一個帳號幾多錢，網軍價格全揭露，天下雜誌671期，2019/04/24。
[25] 林佳賢，跟著資料記者追網軍，「假外國人」如何在PTT鼓吹韓流，天下雜誌671期，2019/04/24。