Android Malware Characterization Using Metadata and Machine Learning Techniques

Table 3

Summary of features, description, and values of the metadata collected from Google Play.

Name		Description	Value
Intrinsic features
1	size	Application size in bytes	Numeric
2	categoryName	Assigned Google Play Category	Categoric
3	ageInMarket	Number of days the app has been on Google Play	Numeric
4	lastSignatureUpdate	Number of days from last app signature update	Numeric
5	timeFromCreation	Number of days since the application was developed	Numeric
6	lastUpdate	Number of days since the application was last updated	Numeric
7	certVal	Number of days from which application is valid	Numeric
8	oldestDateFile	Number of days from the creation of the oldest file in the application	Numeric
9	numPerm	Total number of permissions required by the application	Numeric
10	numFiles	Total number of files the application contains	Numeric
11	numImages	Total number of images the application contains	Numeric
12	numDownloads	Total number of times the application has been uploaded	Numeric
13	versionCode	Google Play reported version of the application	Numeric
14	f+number features	Each of the different Feature hashes	Numeric
Social-related features
15	totalVotes	Total number of rating votes given to the application	Numeric
16	OneStarRatingCont	Number of one-star votes received	Numeric
17	twoStarRatingCont	Number of two-star votes received	Numeric
18	threeStarRatingCont	Number of three-star votes received	Numeric
19	fourStarRatingCont	Number of four-star votes received	Numeric
20	fiveStarRatingCont	Number of five-star votes received	Numeric
21	meanStar	weighted average rating of the application	Numeric
Entity-related features
22	developerRep	Developer reputation metric	Numeric
23	issuerRep	Issuer reputation metric	Numeric
Label
L	isMalware	True if flagged by one or more AV engines	Boolean