隨著網路媒體的普及，用戶生成的內容大量被作為即時資訊來源，但講究即時性的同時也犧牲真實性，有更多未經驗證的錯誤消息被製造出來，嚴重者甚至造成社會動盪、影響經濟發展，因此謠言的控制成為亟需解決的問題。在本篇論文中，我們藉由觀察訊息在推特中的傳遞模式提出驗證謠言真實性的系統，並模擬消息串中動態的信任狀態。為了預測謠言的真實性，我們從訊息與使用者資料提取了幾個謠言的特徵，並結合用卷積神經網絡（CNN）做文字立場分類的結果，再透過隱藏式馬可夫模型（HMM）分別計算其與真假訊息的相似度，生成這些特徵的時間序列的預測結果。特別的是，我們也由模型中取出的狀態序列計算信任度變化。另外，我們採用階層狀架構模擬可信度的傳播，由事件、訊息和回應組成的三層網絡，找出討論相似事件的訊息，以此模擬信任度在社群網路中傳遞的概念，由不同的層級觀察與疊代最佳化產生更好的預測結果。實驗部分涵蓋六個在推特與微博上的中英文真實新聞事件，其中包含謠言與非謠言一共有4,117則訊息與10,484則回應。評估顯示了我們的方法，並與以前的技術相比有14%的改進。可以預測謠言正確性的準確率76％，最後，我們提供真實案例討論在訊息串中信任度變化的情形，以及解釋不同情境下使用此模型的結果，分析優缺點與適用情境，結果顯示利用對話串的內容對與辨認謠言真實性與狀態是有幫助的。

As a source of information, truthfulness of user-generated content is becoming even more important with the prevalence of online media data. In this thesis we develop a system for verification of rumors that propagate through Twitter, and next, we model the dynamic trust states of message threads. To predict the veracity of rumors, we devise several features of rumors and combine with stance features extracted from a convolutional neural network (CNN). Then, the predicted results of a time series of these features are generated using Hidden Markov Models (HMM). The outcome state lists are also used to compute the trust of message thread states which was neglected before. To simulate credibility propagation, a three-layer network consisting of event, sub-events, and messages represents it from a different scale. The verification algorithm was tested on 6 real newsworthy events representing 4,117 posts and 10,484 responses from both Twitter and Weibo. Evaluation demonstrates the efficacy of our approach in comparison with previous state-of-the-art. The system can predict the veracity of rumors with an accuracy of 76%. In addition, real cases are given and discussed to present a better insight of view. The ability to track rumors and predict veracity may help minimize the impact of false information on Twitter.

[1] A. Zubiaga, E. Kochkina, M. Liakata, R. Procter, M. Lukasik, "Stance classification in Rumours as a Sequential Task Exploiting the Tree Structure of Social Media Conversations," Proceedings of the International Conference on Computational Linguistics (COLING). 2016.
[2] Castillo, Carlos, Marcelo Mendoza, and Barbara Poblete, "Information credibility on twitter," Proceedings of the 20th international conference on World wide web. ACM, 2011.
[3] Chen, Yi-Chin, Zhao-Yand Liu, and Hung-Yu Kao, "IKM at SemEval-2017 Task 8: Convolutional Neural Networks for Stance Detection and Rumor Verification," Proceedings of SemEval (2017).
[4] Derczynski, Leon, et al, "SemEval-2017 Task 8: RumourEval: Determining rumour veracity and support for rumours," arXiv preprint arXiv:1704.05972 (2017).
[5] Dong, Xin Luna, et al, "Knowledge-based trust: Estimating the trustworthiness of web sources," Proceedings of the VLDB Endowment 8.9 (2015): 938-949.
[6] Friggeri, Adrien, et al, "Rumor Cascades. ICWSM. 2014.".
[7] H. Abdi and L. J. Williams, "Principal component analysis," Wiley Interdisciplinary Reviews: Computational Statistics, 2:433–459, 2010.
[8] Hamidian, Sardar, and Mona T. Diab, "Rumor Identification and Belief Investigation on Twitter," WASSA@ NAACL-HLT. 2016.
[9] Jin, Zhiwei, et al, "News credibility evaluation on microblog with a hierarchical propagation model," Data Mining (ICDM), 2014 IEEE International Conference on. IEEE, 2014.
[10] Kim., Yoon, "Convolutional neural networks for sentence classification," arXiv preprint arXiv:1408.5882 (2014).
[11] Kreuz, Roger J., and Gina M. Caucci, "Lexical influences on the perception of sarcasm," Proceedings of the Workshop on computational approaches to Figurative Language. Association for Computational Linguistics, 2007.
[12] Kwon, Sejeong, Meeyoung Cha, and Kyomin Jung, "Rumor Detection over Varying Time Windows," PLOS ONE 12.1 (2017): e0168344.
[13] Liu and A. Datta, "Modeling context aware dynamic trust using hidden markov model," In AAAI, pages 1938–1944, 2012.
[14] Liu, Xiaomo, et al, "Real-time rumor debunking on twitter," Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM, 2015.
[15] Lukasik, Michal, Srijith, P.K, Vu, Duy, Bontcheva, Kalina, Zubiaga, Arkaitz and Cohn, "Hawkes processes for continuous time sequence classification: an application to rumor stance classification in twitter," Proceedings of 54th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics, 2016.
[16] M. Gupta, P. Zhao, and J. Han, "Evaluating Event Credibility on Twitter," In Proc. of the 2012 SIAM International Conference on Data Mining(SDM), pp. 153-164. SIAM / Omnipress, 2012.
[17] Mendoza, Marcelo, Barbara Poblete, and Carlos Castillo, "Twitter under crisis: can we trust what we RT?," "Proceedings of the first workshop on social media analytics. ACM, 2010.
[18] Pennington, Jeffrey, Richard Socher, and Christopher D. Manning, "Glove: Global vectors for word representation," EMNLP. Vol. 14. 2014.
[19] Procter, Rob, Farida Vis, and Alex Voss, "Reading the riots on Twitter: methodological innovation for the analysis of big data," International journal of social research methodology 16.3 (2013): 197-214.
[20] Qazvinian, Vahed, et al, "Rumor has it: Identifying misinformation in microblogs," Proceedings of the Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, 2011.
[21] Rabiner, Lawrence R, "A tutorial on hidden Markov models and selected applications in speech recognition," Proceedings of the IEEE 77.2 (1989): 257-286.
[22] Reichel, Uwe D., and Piroska Lendvai, "Veracity computing from lexical cues and perceived certainty trends," arXiv preprint arXiv:1611.02590 (2016).
[23] S. Young, M. Gales, T. Hain, X. Liu, D. Kershaw y G. Evermann, "The HTK Book. Revised for HTK Version 3.4," Cambridge University Engineering Department, 2006.
[24] Szeliski, Richard, "Computer vision: algorithms and applications," Springer Science & Business Media, 2010.
[25] Takahashi, Tetsuro, and Nobuyuki Igata, "Rumor detection on twitter," Soft Computing and Intelligent Systems (SCIS) and 13th International Symposium on Advanced Intelligent Systems (ISIS), 2012 Joint 6th International Conference on. IEEE, 2012.
[26] Wang, Shihan, and Takao Terano, "Detecting rumor patterns in streaming social media," Big Data (Big Data), 2015 IEEE International Conference on. IEEE, 2015.
[27] Wu, Ke, Song Yang, and Kenny Q. Zhu, "False rumors detection on sina weibo by propagation structures," 2015 IEEE 31st International Conference on Data Engineering. IEEE, 2015.
[28] Yang, Fan, et al, "Automatic detection of rumor on Sina Weibo," Proceedings of the ACM SIGKDD Workshop on Mining Data Semantics. ACM, 2012.
[29] Zhao, Zhe, Paul Resnick, and Qiaozhu Mei, "Enquiring minds: Early detection of rumors in social media from enquiry posts," Proceedings of the 24th International Conference on World Wide Web. ACM, 2015.
[30] Zheng, Xiaoming, Yan Wang, and Mehmet A. Orgun, "Modeling the dynamic trust of online service providers using HMM," Web Services (ICWS), 2013 IEEE 20th International Conference on. IEEE, 2013.
[31] Zubiaga, Arkaitz, et al, "Analysing how people orient to and spread rumours in social media by looking at conversational threads," PloS one 11.3 (2016): e0150989.