隨著網際網路在我們生活中蓬勃發展，使用社群媒體已是我們生活不可或缺的一部分，但伴隨而來的是單一使用者註冊多個帳號(稱之為分身帳號)以便在社群媒體進行廣告發文、散播垃圾資訊或是引起爭端等現象，此現象在論壇型社群媒體上更為明顯。識別單一使用者所註冊的多個帳戶是解決此現象的關鍵步驟，然而目前的研究大多是識別跨社群媒體的單一使用者，且其標的多為個人型社群媒體，故本研究提出SiMAIM辨識方法，欲針對一個使用者在一個論壇型社群媒體中註冊的多個分身帳號。我們以台灣論壇型社群媒體Mobile01來驗證SiMAIM，透過蒐集帳號基本資訊、發文內容及社交網絡結構，可以辨識單一使用者所建立的分身帳號。

As the Internet has flourishesd, the use of social media is are indispensable toan integral part of our livesdaily life, . It is observed that a user may own multiple accounts (known as sockpuppets) in a social media website (especially a forum-based one) to advertise products, spread junk information, arouse controversy, etc. but single user registers multiple accounts (called sockpuppets) to spread spam information in social media.This phenomenon is more obvious in the forum-based social media.Identifying multiple accounts registered by a single user is a key step in solving this problem. However, most of the currentexisting research studies is focus on to identifying multiple accounts owned by a single user in cross social mediaacross several social media websites, instead of being in a single website. ThereforeTo address this research gap, our studywe proposes SiMAIM identification method to find a user who has sockpuppets of a user in a forum-type based social media website. We verify the SiMAIM on the data collected from the Mobile01, which is the largest use the Taiwan forum-based social media Mobile01 to verify SiMAIMwebsite in Taiwan. By collecting basic account information, posting content and constructing social network of accounts , we canthe SiMAIM effectively identifiesy sockpuppets created by a single user.

第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究問題與目的 3
第二章 文獻探討 5
2.1 跨社群媒體的使用者身分識別 5
2.2 偵測社群媒體中的垃圾資訊發送者 10
第三章 研究方法 12
3.1 研究架構 12
3.2 資料蒐集 13
3.3 基本資料相關屬性 14
3.4 文章內容相關屬性 15
3.4.1 去除停用詞 16
3.4.2 中文斷詞 16
3.4.3 Term Frequency Inverse Document Frequency 17
3.4.4 LDA 18
3.5 社交網絡相關屬性 20
3.6 辨識分身帳號 25
3.6.1 線性加總 26
3.6.2 分類模型 27
第四章 實驗結果 29
4.1 實驗環境 29
4.2 實驗資料 29
4.3 資料處理 29
4.4 評估指標 32
4.5 實驗結果 32
4.5.1 線性加總實驗結果 33
4.5.1.1 使用線性加總於第一筆測試資料實驗結果 34
4.5.1.2 使用線性加總於第二筆測試資料實驗結果 36
4.5.1.3 使用線性加總於第三筆測試資料實驗結果 38
4.5.2 分類模型實驗結果 40
4.5.2.1 使用分類模型於第一筆測試資料實驗結果 41
4.5.2.2 使用分類模型於第二筆測試資料實驗結果 43
4.5.2.3 使用分類模型於第三筆測試資料實驗結果 45
4.6 比較實驗 48
4.6.1 使用KLD於第一筆測試資料實驗結果 48
4.6.2 使用KLD於第二筆測試資料實驗結果 50
4.6.3 使用KLD於第三筆測試資料實驗結果 51
第五章 結論 54
參考文獻 55

Almishari, M., & Tsudik, G. (2012). Exploring linkability of user reviews. Paper presented at the European Symposium on Research in Computer Security.
Benevenuto, F., Magno, G., Rodrigues, T., & Almeida, V. (2010). Detecting spammers on twitter. Paper presented at the Collaboration, electronic messaging, anti-abuse and spam conference (CEAS).
Benevenuto, F., Rodrigues, T., Almeida, V., Almeida, J., & Gonçalves, M. (2009). Detecting spammers and content promoters in online video social networks. Paper presented at the Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval.
Blei, D. M., Ng, A. Y., & Jordan, M. I. J. J. o. m. L. r. (2003). Latent dirichlet allocation. 3(Jan), 993-1022.
Bron, C., & Kerbosch, J. (1973). Algorithm 457: finding all cliques of an undirected graph. Communications of the ACM, 16(9), 575-577.
Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE: synthetic minority over-sampling technique. Journal of artificial intelligence research, 16, 321-357.
Chen, L.-C. (2017). An effective LDA-based time topic model to improve blog search performance. 53(6), 1299-1319.
Chu, Z., Gianvecchio, S., Wang, H., & Jajodia, S. (2010). Who is tweeting on Twitter: human, bot, or cyborg? Paper presented at the Proceedings of the 26th annual computer security applications conference.
Chung, C. T., Lin, C. J., Lin, C. H., & Cheng, P. J. (2014). Person identification between different online social networks. Paper presented at the Proceedings of the 2014 IEEE/WIC/ACM International Joint Conferences on Web Intelligence (WI) and Intelligent Agent Technologies (IAT)-Volume 01.
Geman, S., Geman, D. J. I. T. o. p. a., & intelligence, m. (1984). Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. (6), 721-741.
Goga, O., Lei, H., Parthasarathi, S. H. K., Friedland, G., Sommer, R., & Teixeira, R. (2013). Exploiting innocuous activity for correlating users across sites. Paper presented at the Proceedings of the 22nd international conference on World Wide Web.
Irani, D., Webb, S., & Pu, C. (2010). Study of Static Classification of Social Spam Profiles in MySpace. Paper presented at the Icwsm.
Jain, P., Kumaraguru, P., & Joshi, A. (2013). @ i seek'fb. me': Identifying users across multiple online social networks. Paper presented at the Proceedings of the 22nd international conference on World Wide Web.
Jain, P., & Kumaraguru, P. J. a. p. a. (2012). Finding Nemo: searching and resolving identities of users across online social networks.
Motoyama, M., & Varghese, G. (2009). I seek you: searching and matching individuals in social networks. Paper presented at the Proceedings of the eleventh international workshop on Web information and data management.
Nie, Y., Jia, Y., Li, S., Zhu, X., Li, A., & Zhou, B. J. N. (2016). Identifying users across social networks based on dynamic core interests. 210, 107-115.
Salton, G., & McGill, M. J. (1986). Introduction to modern information retrieval.
Tan, S., Guan, Z., Cai, D., Qin, X., Bu, J., & Chen, C. (2014). Mapping Users across Networks by Manifold Alignment on Hypergraph. Paper presented at the AAAI.
Vosecky, J., Hong, D., & Shen, V. Y. (2009). User identification across multiple social networks. Paper presented at the Networked Digital Technologies, 2009. NDT'09. First International Conference on.
Xu, H., & Sun, Y. (2015). Identify user variants based on user behavior on social media. Paper presented at the Computing and Communications Conference (IPCCC), 2015 IEEE 34th International Performance.
Zafarani, R., & Liu, H. (2013). Connecting users across social media sites: a behavioral-modeling approach. Paper presented at the Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining.
Zhang, Y., Wang, L., Li, X., & Xiao, C. (2016). Social identity link across incomplete social information sources using anchor link expansion. Paper presented at the Pacific-Asia Conference on Knowledge Discovery and Data Mining.
Zheng, R., Li, J., Chen, H., & Huang, Z. J. J. o. t. A. s. f. i. s. t. (2006). A framework for authorship identification of online messages: Writing‐style features and classification techniques. 57(3), 378-393.
Zhou, X., Liang, X., Zhang, H., & Ma, Y. J. I. t. o. k. d. e. (2016). Cross-platform identification of anonymous identical users in multiple social media networks. 28(2), 411-424.
王力弘. (2015). 社群媒體新詞偵測系統 以 PTT 八卦版為例. 王力弘,
詠勝科技有限公司. mobile01. Retrieved from https://www.mobile01.com/
維基百科編者. 三星寫手門事件. Retrieved from https://zh.wikipedia.org/w/index.php?title=%E4%B8%89%E6%98%9F%E5%AF%AB%E6%89%8B%E9%96%80%E4%BA%8B%E4%BB%B6&oldid=51055105