在科技不斷進步的世代，溝通方式不斷的演變(如:即時通訊，社交網路和聊天等)及廣泛的被採用，而「電子郵件行銷」在企業經營上一直是個重要商業經濟模式，企業可透過電子郵件提供顧客「優惠價格」及「資訊分享」，希望藉此提高與顧客的黏著度，維繫企業與顧客間關係。
電子報發送平台每小時可以寄再多的EMAIL、信件到達率再高，收件者沒有開啟信件，那後續預期的一切效益與計算，也都將隨著這封未開啟的郵件永遠淹沒在郵件大海裡。
本研究希望透過人工智慧分類出收件者開信時間的習慣，因此藉由電子豹公司（NewsLeopard）所提供的寄送時間及開啟信件歷史資料，做為研究數據的基礎，針對歷史資料進行前置處理並將開啟信時段做分類，使用k最近鄰演算法(kNN)建立分類模型，並觀察混淆矩陣(Confusion Matrix)作為評估分類模型的效果，進行不同k值的分類模型比較分析。

In this age of technological improvement, communication methods are evolving constantly (e.g. instant messaging, social media and social messaging, etc.) and are widely used in our everyday lives. ”E-mail marketing” has long been an important business strategy. Businesses can provide customers with “discount offers” and share information through e-mails, so as to increase brand awareness and loyalty and sustain the business-customer relationship.
E-newsletter platforms can send as many e-mails per hour, and perhaps even have a high rate of successfully sent emails. However, if the receiving end never opens the mail, all expected benefits and calculations that follows would be lost.
By categorizing the habits of newsletter receivers when opening e-newsletters with artificial intelligence, this study will analyze the data on sending time and history of e-mails opened as provided by NewsLeopard as basis. The study will pre-process such data and categorize according to the opening time of e-mails, thereby building a categorization model with kNN. Confusion matrix will be used to evaluate the efficiency of the categorization model by comparing the model with different K-values.

【中文文獻】
[1]傅馨巧(2005)，「電子郵件行銷之研究人們為什麼閱讀及轉信」，東吳大學企業管理研究所碩士論文
[2]謝如婷(2017)，「電子郵件行銷發送策略與開信率之研究」，國立政治大學商學院經營管理碩士學程國際金融組碩士論文
[3]李文慈(2005) ，「以監督式叢集分析方式使用基因表現資料尋找癌症子型」，長庚大學資訊管理研究所碩士論文
[4]Python機器學習(博碩文化)
[5]古祐嘉(2009)，「適應性K最近鄰演算法」，元智大學資訊管理學系碩士班
[6]林嘉陞(2007)，「CANN：一個整合分群中心與最鄰近鄰居之入侵偵測系統」，國立中正大學會計與資訊科技研究所碩士論文
[7]陳柏均(2011)，文件距離為基礎kNN分群技術與新聞事件偵測追蹤之研究，政治大學資訊管理研究所碩士論文
[8]吳季桓(2000)，「自動分類的實作：KNN 與 SVM」，國立中正大學資訊工程研究所碩士論文
[9]戴文惠(2003)，「僑鵬旅行社e-mail行銷策略之研究」，東吳大學企業管理研究所碩士論文
[10]林鼎家(2013)，「比較特徵選擇與特徵擷取方法在評估中文適讀性的效果」，嘉義大學理工學院資訊工程學系碩士論文
[11]張凱評(2016) ，「應用資料探勘技術於電話行銷成功與否之預測」，淡江大學資訊工程學系碩士在職專班碩士論文
[12]林哲宏(2008) ，「影響電子報開信率、點擊率與轉寄行為之因素探討-以台灣某知名傢俱居家修繕中心為例」，國立臺北科技大學工業工程與管理研究所碩士論文
[13]陳逸真(2017)，「不平衡資料分類方法之比較」，國立臺北大學統計學系碩士論文
自動分類的實作：KNN與SVM」，國立中正大學資訊工程研究所碩士論文
[14]黃國祥(2005)，「網路商店應用電子郵件行銷策略之研究」，靜宜大學資訊管理學系研究所碩士論文
[15]吳翰群(2014) ，「運用資料探勘技術於廣告點擊率預測」，國立交通大學管理學院資訊管理學程碩士論文
[16]吳俊穎(2013)，「利用相似比例加權改善KNN演算法於偏斜資料之分類效能」，中華大學資訊工程學系碩士班碩士論文

【英文文獻】
[17]Wenqian Shang, Haibin Zhu, Houkuan Huang, Youli Qu, and Yongmin Lin , "The Improved Ontology kNN Algorithm and its Application," Proceedings of the 2006 IEEE International Conference on Networking, Sensing and Control, pp. 198 - 203, 2006.
[18]Chunming Liu, Longbing Cao, Philip S Yu, "A hybrid coupled k-nearest neighbor algorithm on imbalance data," Proceedings of the 2014 International Joint Conference on Neural Networks (IJCNN), pp. 2011 – 2018, 2014.
[19]Dawei Yin, Chang An, Henry S. Baird , "Imbalance and Concentration in k-NN Classification," Proceedings of the 2010 20th International Conference on Pattern Recognition, pp. 2170 – 2173, 2010.
[20]Okfalisa, Ikbal Gazalba, Mustakim, Nurul Gayatri Indah Reza , "Comparative analysis of k-nearest neighbor and modified k-nearest neighbor algorithm for data classification," Proceedings of the 2017 2nd International conferences on Information Technology, Information Systems and Electrical Engineering (ICITISEE), pp. 294 – 298, 2017.
[21]Shiliang Sun, Rongqing Huang , "An adaptive k-nearest neighbor algorithm," Proceedings of the 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery, pp. 91 – 94, 2010.
[22]I-Ling Chen, Kai-Chih Pai, Bor-Chen Kuo, Cheng-Hsuan Li , "An Adaptive Rule Based on Unknown Pattern for Improving K-Nearest Neighbor Classifier," Proceedings of the 2010 International Conference on Technologies and Applications of Artificial Intelligence, pp. 331 – 334, 2010.
[23]Zeng Yong, Wang Bing, Zhao Liang, Yang Yupu , "The extended nearest neighbor classification," Proceedings of the 2008 27th Chinese Control Conference, pp. 559 – 563, 2008.
[24]Bing Wang, Yong Zeng, Yupu Yang , "Generalized nearest neighbor rule for pattern classification," Proceedings of the 2008 7th World Congress on Intelligent Control and Automation, pp. 8465 – 8470, 2008.
[25]Tung-Shou Chen, Yung-Hsing Chiu, Chih-Chiang Lin , "Fast Nearest Neighbor Classification using Class-Based Clustering," Proceedings of the 2007 International Conference on Machine Learning and Cybernetics, pp. 1894 – 1898, 2007.
[26]Kaibo Zhou, Gaofeng Xu, Sihai Guo , "Anode effect prediction based on support vector machine and K nearest neighbor," Proceedings of the 2017 Chinese Automation Congress (CAC), pp. 341 - 345, 2017.
[27]J. Laaksonen, E. Oja , "Classification with learning k-nearest neighbors," Proceedings of the Proceedings of International Conference on Neural Networks (ICNN'96), pp. 1480 - 1483 vol.3, 1996.
[28]Naohiro Ishii, Ippei Torii, Yongguang Bao, Hidekazu Tanaka , "Mapping of nearest neighbor for classification," Proceedings of the 2013 IEEE/ACIS 12th International Conference on Computer and Information Science (ICIS), pp. 121 – 126, 2013.
[29]Yu-Long Qiao, Jeng-Shyang Pan, Sheng-He Sun , "Improved K nearest neighbor classification algorithm," Proceedings of the The 2004 IEEE Asia-Pacific Conference on Circuits and Systems, 2004. Proceedings, pp. 1101 - 1104 vol.2, 2004.
[30]Horng-Lin Shieh , "Semi-supervised Clustering Based on K-Nearest Neighbors," Proceedings of the 2012 Third International Conference on Digital Manufacturing & Automation, pp. 759 - 762, 2012.
【網站資料】
[31]Email Statistics Report, 2017-2021 (Radicati報告)
http://www.radicati.com/wp/wp-content/uploads/2017/01/Email-Statistics-Report-2017-2021-Executive-Summary.pdf
[32]MailChimp 調查
https://mailchimp.com/resources/research/email-marketing-benchmarks/
[33]數位時代：影響收件者開信的三大因素
https://www.bnext.com.tw/ext_rss/view/id/668153
[34]翻轉人類未來的 AI 科技：機器學習與深度學習
https://technews.tw/2017/10/05/ai-machine-learning-and-deep-learning/
[35]人工類神經網路
http://terms.naer.edu.tw/detail/1679037/