The Study of Dynamic Analysis Mechanisms for Android Malware
| Autor: | Zhen-Gao Liu, 劉朕誥 |
|---|---|
| Rok vydání: | 2015 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 103 Today, smart phones are everywhere and changing our daily life because of the universal of smart phones and the mobile Internet. However, most people know few about the security of smart phones, which makes users install a malware that would leak privacy information inadvertently. Android, one of the most popular system, is convenient for exploitation because of its property of open source, but that makes it become an attack target for malwares easily. Although anti-virus software is in progress continuously, it is not enough for daily-changing malwares. Our purpose of this thesis is to provide a system which can automatically detect malwares by Droidbox, a dynamic analysis tool, with Docker container technology dynamically. We use support vector machine (SVM), which is better in accuracy, to detect whether the same type applications are malwares or not by training and building models from a large number of same type application features of benign applications and malicious applications. The result shows that our proposed system has a quite high positive rate and low false rate when detecting malware in the same type applications. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|