講演抄録/キーワード |
講演名 |
2017-11-15 13:50
Machine Learning Approach for Phishing Detection in SDN Networking ○Yu-Hung Chen・Jiun-Yu Yang・Po-Chun Hou・Jiann-Liang Chen(National Taiwan University of Science & Technology) IA2017-30 |
抄録 |
(和) |
People have become increasingly dependent on information technology since the emergence of the Internet. Therefore, many hackers engage in financial crimes and computer attacks through the Internet. The existing attack modes include a combination of Trojans, Botnet, social engineering, and email phishing technology. In recent years, the rise of numerous phishing attacks has led to great money loss on the part of people deceived by phishing attacks. In general, phishing websites have a short lifetime, and the attack technique is relatively more complex, thus drawing information security concern. In this study, the software-defined infrastructure was adopted to deploy a platform where automated phishing website risk analysis, assessment, and strategic executions and operations were conducted. During the experimentation period, 25,000 phishing websites were obtained using Common Crawl, and 25,000 illegal websites were obtained from Phishing Tank. Machine learning was conjunctively used in this study to complete the research and development of the phishing website learning model creation, risk analysis, hazard assessment, and other mechanisms. Software-defined networking was subsequently used to carry out strategic execution targeting hazardous messages in order to ensure a high degree of overall network environment security. The experimental results show that the training model in this study reaches an accuracy of 96.97%. |
(英) |
People have become increasingly dependent on information technology since the emergence of the Internet. Therefore, many hackers engage in financial crimes and computer attacks through the Internet. The existing attack modes include a combination of Trojans, Botnet, social engineering, and email phishing technology. In recent years, the rise of numerous phishing attacks has led to great money loss on the part of people deceived by phishing attacks. In general, phishing websites have a short lifetime, and the attack technique is relatively more complex, thus drawing information security concern. In this study, the software-defined infrastructure was adopted to deploy a platform where automated phishing website risk analysis, assessment, and strategic executions and operations were conducted. During the experimentation period, 25,000 phishing websites were obtained using Common Crawl, and 25,000 illegal websites were obtained from Phishing Tank. Machine learning was conjunctively used in this study to complete the research and development of the phishing website learning model creation, risk analysis, hazard assessment, and other mechanisms. Software-defined networking was subsequently used to carry out strategic execution targeting hazardous messages in order to ensure a high degree of overall network environment security. The experimental results show that the training model in this study reaches an accuracy of 96.97%. |
キーワード |
(和) |
Phishing detection / Software-Defined Networking / Extreme Learning Machine / Ensemble Learning / Feature engineering / / / |
(英) |
Phishing detection / Software-Defined Networking / Extreme Learning Machine / Ensemble Learning / Feature engineering / / / |
文献情報 |
信学技報, vol. 117, no. 299, IA2017-30, pp. 1-6, 2017年11月. |
資料番号 |
IA2017-30 |
発行日 |
2017-11-08 (IA) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
IA2017-30 |
|