METADroid: lightweight Android classification using meta-data

dc.contributor.advisorStakhanova, Natalia
dc.contributor.authorLi, Yan
dc.date.accessioned2023-03-01T16:21:10Z
dc.date.available2023-03-01T16:21:10Z
dc.date.issued2016
dc.date.updated2020-11-25T00:00:00Z
dc.description.abstractThe Android system is the most widely used mobile system in the world and the user requirement is still increasing. According to [37], Android dominated the market with a 79.8% in 2013 Q2, 84.8% in 2014 Q2 and 82.8% share in 2015 Q2. Compared to other mobile systems, Android dominates most of the market and has the largest number of mobile users. Based on the research work [44] from PulseSecure.net, the number of new Android malware samples have been dramatically increased from 3809 in 2011, 214327 in 2012, 1192035 in 2013, 1548129 in 2014 and in the 2015 Q1 the new malicious sample is 440267. The approximate total value during the entire year of 2015 was a greater total than 2014. With that in mind, malware has always been the most pressing concern for the mobile application market. There are a number of analysis tools and architectures used for malware detection including static analysis, dynamic analysis, sandbox analysis and manually dissection. Consequently, all of the analysis approaches are time consuming. In order to effectively use our limited time and human resources, we present a lightweight pre-filtering tool(METADroid) which could be used to pre-classify the apps before the more expensive traditional static, dynamic analysis. The system includes whole 381 features and we analyzed their effectiveness for triaging. It will give you feedback of each apk in a short period of time and provide valuable prediction. Our experiments based on more than 158000 Android Applications collected from 8 markets around the globe.
dc.description.copyright© Yan Li, 2016
dc.formattext/xml
dc.format.extentxi, 140 pages
dc.format.mediumelectronic
dc.identifier.otherThesis 9884
dc.identifier.urihttps://unbscholar.lib.unb.ca/handle/1882/13610
dc.language.isoen_CA
dc.publisherUniversity of New Brunswick
dc.rightshttp://purl.org/coar/access_right/c_abf2
dc.subject.disciplineComputer Science
dc.titleMETADroid: lightweight Android classification using meta-data
dc.typemaster thesis
thesis.degree.disciplineComputer Science
thesis.degree.fullnameMaster of Computer Science
thesis.degree.grantorUniversity of New Brunswick
thesis.degree.levelmasters
thesis.degree.nameM.C.S.

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
item.pdf
Size:
2.3 MB
Format:
Adobe Portable Document Format