新聞讀者會持續的從各種管道獲取新聞，藉此了解最新的社會情勢。獲取新聞的管道逐漸的從報紙、新聞媒體以及新聞網站上直接獲取，轉移到從各種不同的社群網絡上取得。社群網絡開始在新聞消費中扮演重要角色，因此社群網絡上的假新聞議題也顯得愈發重要。在社群網絡上由於其本身的特性、大量的惡意帳號等等原因，使得假新聞的傳播非常的快速、廣泛且難以阻擋。這些惡意創造的假新聞，很難由個人辨別，進而在接觸、閱讀到假新聞之後，導致讀者有了錯誤的認知。當這些新聞讀者相信了假新聞中的錯誤資訊，並據此行動及決策，就有可能對個人以及整個社會造成危害。更進一步的來說，當人們意識到假新聞的存在，但是卻又無力辨別時，就會懷疑所有接觸到的新聞及資訊，進而連資訊正確與否都難以辨別。在這樣的狀況下，幫助新聞讀者辨別包含錯誤資訊的新聞，使讀者能夠在開始閱讀之前就意識到潛在的威脅並進而能夠避免受到假新聞的危害，甚至進一步的阻止假新聞的散播即是非常重要的目標。本論文透過新聞是由人事時地物等事實來進行描述這樣的觀察，提出了一個藉由文字模型差異來辨別假新聞的方式。在單一新聞的文字模型建構完成後，就能夠將其與信任的新聞進行比較，根據比較的結果可以辨識出該新聞是否值得信任。當一個新聞來源持續的藉由比較差異被確認為值得信任，那就可以擴充信任的新聞來源，完成更好的辨識結果。

News consumers continue to get news from a variety of channels to learn about the latest situations in the world.
In the current news consumption environment, news consumers access news through social media rather than directly from a news website or newspaper.
Social networks have begun to play an important role in news consumption, so the issue of fake news on social networks is becoming more important.
Due to the sharing nature of the social network and a large number of malicious accounts, the spread of fake news is very fast, extensive and difficult to block.
Individual news readers are difficult to identify fake news on their own and have a wrong belief after accessing and reading the fake news.
When news readers make decisions based on disinformation, it may cause harm to individuals and society.
After people realize the existence of fake news, but they are unable to identify it, they will doubt all the news and unable to judge the correctness of news.
Under such circumstances, it is an important goal to help news consumers to identify news that contains disinformation, so that consumers can realize the potential threats before they start reading and thus avoid being harmed by fake news, and even further prevent the spread of fake news.
This thesis uses the observation that news is constructed with who, when, what, why and how, and proposes a way to distinguish fake news by Text Comparison Model.
After the construction of the text model, it can be compared with trustworthy news. Based on the comparison results, news can be identified whether it is trustworthy or not. When a news source is continuously identified as trustworthy by comparing differences, it is possible to expand the source of trusted news and achieve better identification results.

[1] Hunt Allcott and Matthew Gentzkow. Social media and fake news in the 2016 election. Journal of Economic Perspectives, 31(2):211–36, May 2017.
[2] Ramy Baly, Georgi Karadzhov, Dimitar Alexandrov, James R. Glass, and Preslav Nakov. Predicting factuality of reporting and bias of news media sources. CoRR, abs/1810.01765, 2018.
[3] Zi Chu, Steven Gianvecchio, Haining Wang, and Sushil Jajodia. Who is tweeting on twitter: Human, bot, or cyborg? In Proceedings of the 26th Annual Computer Security Applications Conference, ACSAC ’10, pages 21–30, New York, NY, USA, 2010. ACM.
[4] Giovanni Luca Ciampaglia, Prashant Shiralkar, Luis M. Rocha, Johan Bollen, Filippo Menczer, and Alessandro Flammini. Computational fact checking from knowledge networks. PLOS ONE, 10(6):1–13, 06 2015.
[5] Benjamin D. Horne, William Dron, Sara Khedr, and Sibel Adali. Assessing the news landscape: A multi-module toolkit for evaluating the credibility of news. In Companion Proceedings of the The Web Conference 2018, WWW ’18, pages 235–238, Republic and Canton of Geneva, Switzerland, 2018. International World Wide Web Conferences Steering Committee.
[6] Nitin Jindal and Bing Liu. Opinion spam and analysis. In Proceedings of the 2008 International Conference on Web Search and Data Mining, WSDM ’08, pages 219–230, New York, NY, USA, 2008. ACM.
[7] Quoc V. Le and Tomas Mikolov. Distributed representations of sentences and documents. CoRR, abs/1405.4053, 2014.
[8] Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg Corrado, and Jeffrey Dean. Distributed representations of words and phrases and their compositionality. In Proceedings of the 26th International Conference on Neural Information Processing Systems - Volume 2, NIPS’13, pages 3111–3119, USA, 2013. Curran Associates Inc.
[9] Nic Newman, Richard Fletcher, Antonis Kalogeropoulos, David A. L. Levy, and Rasmus Kleis Nielsen. Reuters Institute Digital News Report 2018, 2018. http://media.digitalnewsreport.org/wp-content/uploads/2018/06/digital-newsreport-2018.pdf Accessed September 2019.
[10] Jeppe Norregaard, Benjamin D. Horne, and Sibel Adali. NELA-GT-2018: A large multi-labelled news dataset for the study of misinformation in news articles. CoRR, abs/1904.01546, 2019.
[11] Zizi Papacharissi and Maria de Fatima Oliveira. Affective news and networked publics: The rhythms of news storytelling on #egypt. Journal of Communication, 62(2):266–282, 02 2012.
[12] Craig Silverman. This Analysis Shows How Viral Fake Election News Stories Outperformed Real News On Facebook. https://www.buzzfeednews.com/article/craigsilverman/viral-fake-election-newsoutperformed-real-news-on-facebook Accessed September 2019.