友善、方便的網路使用環境與嚴格控管的網路使用環境一直是網路管理人員經常面對的兩難問題；控管嚴格的網路使用時一定要經過認證或是事先經過註冊，但對於臨時使用者、外來的訪客，突然或經常需要使用網路的新設備裝置的使用者，就變成是個困擾。如建構一個對網路裝置友善開放易用的網路環境，讓使用者只要插上網路線或自動搜尋無線網路即可使用網路，無須認證，雖然方便卻也同時也帶來管理上的困擾，當使用者不當使用網路或系統被入侵而造成網路的癱瘓時，也不易追蹤找出發生問題的使用者裝置。
在此研究的目標是以S大學學院內的網路環境為範例，透過簡易網路管理協定收集網路裝置的網路卡位址(MAC address)等相關資料，這些裝置可能是一般電腦、筆記型電腦、平板電腦、手機、印表機或透過網路控制的資訊設備。但裝置資料的維護會因為人員設備的異動、地點轉移需要時常更新資料，常造成設備資料缺少的狀況。因此希望能夠過資料探勘的方式，透過現有資料中可靠度高的裝置資訊找出與未知裝置間的關連性，讓經驗不足的管理人員也可以透過資訊系統提供的知識來解決問題。
實驗結果顯示，此方法可有效地識別清晰的標識設備和其他設備都是高度相關的問題，透過本研究的方法可找出問題裝置的接近群組 ，探勘適合人工查核有問題的實體網路位址，數量越少裝置相關性越高。對於網絡管理設備發生問題時有很大的幫助。

A friendly, convenience and security are two contradictory goals faced by network managers. When using a strictly controlled network, users must be certified or registered in advance in order to ensure security. However, for temporary users, outsiders, and people who frequently use new devices, the process is often considered troublesome. An open and friendly internet environment allows users to simply plug-in cables or automatically search with wireless and get on internet without authentication yet compromises internet security. In the later environment, when the network is paralyzed by improper use of users or attacks on system, it is not easy to identify the users of the device that causes the problem.
This study takes College network environment of University S as an example and collect MAC address and relative data of various devices through SNMP (Simple Network Management Protocol). These devices may be personal computers, laptop computers, tablet computers, mobile phones, printers or other devices that are controlled through internet. Due to the change of personnel, equipment and their location, the data about a device sometimes is inaccurate and need to be constantly updated. Therefore, In this thesis, I proposes a data mining approach to identifies devices with reliable data in connection with some problematic device whose data is either missing or inaccurate. These related devices may reveal the real identity of the owner of the problematic device.
Is experimental results show identifying clear identification and other equipment are highly related issues. Through the method of this research may identify problems with the device close group. Mining suitable for physical network address manually check in question, the fewer the number of devices more relevant. For network management equipment problems, there is a great help.

論文審定書 i
誌謝 ii
中文摘要 iii
Abstract iv
表次 vii
圖次 viii
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機 1
1.3 研究目的 2
第二章 文獻探討 4
2.1 網路管理的方式 4
2.2 簡單網路管理協定 5
2.3 位址解析協定 6
2.4 資料探勘 6
2.4.1關聯規則 8
2.4.2 關聯規則的定義 9
2.4.3 關連分析演算法 10
2.5 視覺化呈現 11
第三章 研究方法 14
3.1 網路環境 14
3.2 系統架構 15
3.3 資料收集和清理 16
3.4 資料變數定義 22
3.5 找出接近群組 26
第四章 效能評估 33
4.1 樣本描述 33
4.2 評估效能 35
第五章 系統展示與案例 44
5.1案例一：濫發DHCP的無線網路Access Point 49
5.2 案例二：研究室的研究生架設具有安全漏洞的網路伺服器 51
第六章 結論與未來研究方向 54
中文文獻 55
英文文獻 56

中文文獻
1. Pang-Ning Tan, Michael Steinback, Vipin Kumar, 資料探勘, 台北市: 台灣培生教育, 2007.
2. 李光偉, “學生宿舍網路與管理系統,” 2010. [線上]. Available: http://www.cc.ntu.edu.tw/chinese/epaper/0013/20100620_1301.htm.
3. 謝昇晃, “跨層之網路管理系統：以輔仁大學為實例,” 輔仁大學電子工程學系碩士論文, 2009.
4. 潘立人, “資料搜尋系統視覺化與多維度分析之設計：以資訊工程研究論文檢索系統為例,” 國立中央大學資訊工程學系碩士論文, 2010.
5. 劉興憲, “校園網路管理與整合模式之研究-以國立彰化師範大學宿舍網路為例,” 國立彰化師範大學資訊管理學系碩士論文, 2010.
6. 賴意淳, “具自調式能力之網路流量分析系統,” 銘傳大學資訊傳播工程學系碩士論文, 2009.
7. Douglas Mauro & Kevin Schmidt, “網路管理與S N M P,” 於 SNMP 網路管理實務 第二版, O'Reilly Media, Inc. Taiwan Branch, 2007.
8. 張展華&包蒼龍, “旅館網路系統管理策略之研究,” 於 Workshop on Computer Networks and Web Service/Technologies, 2009.

英文文獻
1. Agrawal, R., Imielinski, T., and Swami, A., "Mining Association Rules between Sets of Items in Large Databases," in 1993 ACM SIGMOD Conference, Washington DC, 1993.
2. B. T. LLC, "BRIDGE-MIB," ByteSphere Technologies LLC, 2006. [Online]. Available: http://www.oidview.com/mibs/0/BRIDGE-MIB.html.
3. Y. Wang, E.-P. Lim, S.-Y. Hwang, "Efficient Mining of Group Patterns from User Movement Data," Data and Knowledge Engineering (DKE), vol. 3, no. 57, pp. 240-282, 2006.
4. Rithme, "Knowledge Discovery and Data Mining," 2014. [Online]. Available: http://www.rithme.eu/?m=resources&p=kdprocess&lang=en.
5. Christian Borgelt, "Find Frequent Item Sets with the FP-growth Algorithm," 2013. [Online]. Available: http://www.borgelt.net/doc/fpgrowth/fpgrowth.html.
6. Agrawal, R., & Srikant, R., “Fast algorithms for mining association rules.,” in Proc. 20th Int. Conf., 1994.
7. Christian Borgelt, "FP-growth - Frequent Item Set Mining," 2014. [Online]. Available: http://borgelt.net/fpgrowth.html.
8. W. Nicholas, “RFC1213-MIB MIB Download,” ByteSphere LLC, and Nicholas W. Saparoff., 2012. [Online]. Available: http://www.oidview.com/mibs/0/RFC1213-MIB.html.
9. GEXF Working Group, "GEXF File Format," Gephi Community project, 2007. [Online]. Available: http://gexf.net/format/.
10. A. JACOMY, "SIGMAJS," 2014. [Online]. Available: http://sigmajs.org.
11. S. Waldbusser, "SNMPWALK," Wes Hardaker, 2002. [Online]. Available: http://www.net-snmp.org/docs/man/snmpwalk.html.
12. Han, J., Pei, J., & Yin, Y., "Mining frequent patterns without candidate generation.," ACM SIGMOD Record, vol. 29, no. 2, pp. 1-12, May 2000.
13. Christian Borgelt, "An Implementation of the FP-growth Algorithm," Proceedings of the 1st international workshop on open source data mining: frequent pattern mining implementations., pp. 1-5, 8 2005.
14. Christian Borgelt, "Frequent item set mining," DATA MINING AND KNOWLEDGE DISCOVERY, vol. 2, p. 437–456, 18 Oct 2012.