本研究主要探討基於社交聊天訊息文本的身份鑑別，近年來，線上社交詐騙行為頻傳，大多情況為利用社交工程手法進行個人帳號之盜用。鑑於此，本研究以此現象為研究標的，希望能建立一套有效率的身分鑑別系統以辨別使用者身份的真實性與合法性。研究方法中將以使用者的社交文本訊息作為使用者鑑別資料來源，並利用潛在語意分析模型(Latent Semantic Analysis; LSA)、多層感知器(Multilayer Perceptron; MLP)與支援向量機(Support Vector Machine; SVM)做為主要的資料分析演算法，進行使用者鑑別符元的產生與鑑別準確率檢測。研究成果顯示，在語意模型分析實驗中，高達65%的檢測案例中，相似度皆低於70%。再者，多層感知器分析與支援向量機分析則分別可達到90%與93%的鑑別準確率。

Recently, cases of scamming on social media keep pouring in. Most cases are related to hacked social media accounts, which belong to those who suffered from identity stealing by social engineering. In this research, we focus on how instant messages of users’ defeat identity thieves. We purposed an authentication system based on users’ instant messages, which is able to tell whether the current user of the account having both of its representation and perpetuity. We collect users’ instant message as the training sets, create the model by utilizing both Latent Semantic Analysis Model and Multilayer Perceptron (MLP) and use Support Vector Machine (SVM) as the classifier. The research result pointed out that, with only Semantic Analysis Model equipped, 65% of test cases cannot reach 70% of similarity, while utilizing Multilayer Perceptron and Support Vector Machine can reach both 90% and 93% of accuracy.

第一章 緒論 1
1.1研究背景與動機 1
1.2研究目的 3
1.3論文架構 3
第二章 文獻探討 4
2.1社群網站 4
2.2 即時通訊 5
2.3 語意模型 6
2.4機器學習 7
2.5身份鑑別 8
第三章 研究方法 10
3.1研究流程 10
3.2資料蒐集 11
3.3正規化輸入資料 12
3.4矩陣處理 12
3.4.1 Term by Document Matrix 12
3.4.2 Term Frequency Inverse Document Frequency 13
3.5語意模型 13
3.5.1潛在語意分析 13
3.6餘弦相似性 14
3.7支援向量機 15
3.8多層感知器 16
第四章 實驗結果 17
4.1 研究環境架設 17
4.2研究資料 17
4.3評估指標 20
4.4資料處理方式與流程 20
4.5 研究結果 23
第五章 結論及未來研究方向 32
5.1 結論與未來研究方向 32
第六章 參考文獻 33

[1] A. Abbasi, H.-C. Chen, “Writeprints: A Stylometric Approach to Identity-Level Identification and Similarity Detection in Cyberspace” ACM Transactions on Information Systems, Vol. 26, Issue 2, No. 7, 2008, DOI: 10.1145/1344411.1344413.
[2] J. Albadarneh, B. Talafha, M. Al-Ayyoub, B. Zaqaibeh, M. Al-Smadi, Y. Jararweh, E. Benkhelifa, “Using Big Data Analytics For Authorship Authentication of Arabic Tweets” In Proceeding of the 8th International Conference on Utility and IEEE Cloud Computing, Limassol, Cyprus, Dec 7-10, 2015.
[3] S. Barbon Jr, R. A. Igawa, B. B. Zarpelão, “Authorship verification applied to detection of compromised accounts on online social networks” Multimedia Tools and Applications, Vol 76, Issue 3, pp. 3213-3233, 2017.
[4] J. Botelho, C. Antunes, “Combining Social Network Analysis with Semi-supervised Clustering: a case study on fraud detection ” In Proceeding of the Mining Data Semantics (MDS'2011) in conjuction with SIGKDD2011, San Diego, CA, Aug. 21-24, 2011.
[5] A. Buriro, B. Crispo, F. Delfrari, “Hold and Sign: A Novel Behavioral Biometrics for Smartphone User Authentication” IEEE Security and Privacy Workshops (SPW), May 22-26, San Jose, CA, USA, 2016.
[6] D. M. Boyd, N. B. Ellison, “Social Network Sites: Definition, History, and Scholarship” Journal of Computer-Mediated Communication, Vol 13, Issue 1, pp. 210-230, 2007.
[7] M. L. Brocardo, I. Traore, I. Woungang, “Authorship verification of e-mail and tweet messages applied for continuous authentication” ACM Journal of Computer and System Sciences, Vol. 81, Issue 8, pp. 1429-1440, 2015.
[8] S. Deerwester, S. T. Dumais, G. W. Furnas, T. K. Landauer, R. Harshman, ”Indexing by latent semantic analysis” Journal of the American Society for Information Science, Vol. 41, Issue 6, pp. 391-407, 1990.
[9] N. E. Evangelopoulos, “Latent semantic analysis” Journal of the Wiley Interdisciplinary Reviews: Cognitive Science, Vol. 4, No. 6, pp. 683-692, 2013.
[10] G. H. Golub, C. Reinsch, “Singular value decomposition and least squares solutions” Journal of the Numerische Mathematik, Vol. 14, No. 5, pp. 403-420, 1970.
[11] P. Gonçalves, M. Araújo, F. Benevenuto, M. Cha, “Comparing and Combining Sentiment Analysis Methods” In Proceedings of the first ACM conference on Online social networks, Oct. 07-08, pp. 27-38, Boston, Massachusetts, USA, 2013.
[12] S. J. Kang, S. Y. Lee, H. I. Cho, “ECG Authentication System Design Based on Signal Analysis in Mobile and Wearable Devices” IEEE Signal Processing Letters, Vol.23, Issue 6, pp. 805-808, Feb. 2016.
[13] R. Klein, A. Kyrilov, M. Tokman, “Automated assessment of short free-text responses in computer science using latent semantic analysis” In Proceedings of the Sixteenth Annual Joint Conference on Innovation and Technology in Computer Science Education (ITiCES 2011), June 27-29, pp. 158-162, Darmstadt, Germany, 2011.
[14] F.-F. Kuo, M.-K. Shan, S.-Y. Lee, “Background music recommendation for video based on multimodal latent semantic analysis” In Proceedings of the IEEE International Conference on Multimedia and Expo , IEEE-ICME 2013, July 15-19, pp. 1-6, San Jose, CA, USA, 2013.
[15] T. K. Landauer, D. S. McNamara, S. Dennis, W. Kintsch, Handbook of Latent Semantic Analysis, Psychology Press, London, UK, 2013.
[16] C. D. Manning, P. Raghavan, H. Schütze, Introduction to information retrieval, Cambridge University Press Cambridge, 2008.
[17] J. Mantyjarvi, J. Himberg, T. Seppanen “Recognizing human motion with multiple acceleration sensors” In Proceeding of the 2001 IEEE International Conference on Systems, Man, and Cybernetics, Oct. 7-10, Tucson, AZ, USA, USA, 2001.
[18] B. A. Nardi, S. Whittaker, E. Bradner, “Interaction and Outeraction: Instant Messaging in Action” In Proceedings of the 2000 ACM conference on Computer supported cooperative work (CSCW '00), Dec. 02-06, pp.79-88, Philadelphia, Pennsylvania, USA , 2000.
[19] M. G. Ozsoy, F. N. Alpaslan, I. Cicekli, “Text summarization using latent semantic analysis” Information Science, Vol. 37, Issue 4, pp. 405-417, 2011.
[20] F. Rosenblatt, “The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain” Psychological Review, pp. 386-408, 1958.
[21] A. L. Samuel, “Some studies in machine learning using the game of checkers” IBM Journal of Research and Development, Vol. 3, Issue 3, pp. 210-219, 1959.
[22] V. Vapnik, C. Cortes, “Support-Vector Networks” Journal of Machine Learning, Vol.20, Issue 3, pp. 273-297, 1995.
[23] S.-H. Wu, M.-J. Chou, C.-H. Tseng, “Detecting In Situ Identity Fraud on Social Network Services: A Case Study With Facebook” IEEE Systems Journal, Vol. 11, Issue 4, pp. 2432-2443, 2017.
[24] R. Zheng, J. Li, H. Chen, Z. Huang, “A Framework for Authorship Identification of Online Messages: Writing-Style Features and Classification Techniques” Journal of the American Society for Information Science and Technology, Vol.57, Issue 3, pp. 378-393, 2006.
[25] Vincenzo Cosenza, http://vincos.it/world-map-of-social-networks/2017 (accessed on 15th April 2018)
[26] 八成以上台灣人愛用Facebook、Line坐穩社群網站龍頭 1人平均擁4個社群帳號年輕人更愛YouTube和IG https://www.iii.org.tw/Press/NewsDtl.aspx?nsp_sqno=1934&fm_sqno=14 (accessed on 15th April 2018)
[27] 台灣活躍用戶破1800萬人，Facebook鎖定電商發力https://www.bnext.com.tw/article/40252/BN-2016-07-19-174028-223 (accessed on 15th April 2018)
[28] 盜用LINE帳號誆稱借錢 中老年族群最易被騙http://www.chinatimes.com/realtimenews/20170415004271-260402 (accessed on 15th April 2018)
[29] 資策會FIND/經濟部技術處「資策會FIND(2016)/ 服務系統體系驅動新興事業研發計畫（2/4）」，https://www.iii.org.tw/Press/NewsDtl.aspx?fm_sqno=14&nsp_sqno=1952 (accessed on 15th April 2018)
[30] 社群新寵兒:即時通訊軟體全球使用率上升12%，更多網路使用者選擇非開放的社群平台 http://www.cna.com.tw/postwrite/Detail/179665.aspx#.WkDY9t-WZhF (accessed on 15th April 2018)
[31] 維基百科 https://zh.wikipedia.org/wiki/身份验证 (accessed on 15th April 2018)
[32] 維基百科 https://zh.wikipedia.org/wiki/助詞 (accessed on 15th April 2018)