Masoomeh Beitsayahi - Yadegar-e-Imam Khomeini (RAH) Shahre Rey Branch, Islamic Azad University, Tehran, Iran

Today, due to the increase in malware and the level of access to information onmobile phones, it is necessary to identify the types of malware. In this research we try toidentify Android malware, at first collected ۱۸۰۴ unsupervised data from Google Play andthird-party markets. And after initial reverse engineering and decompiling them and identifyingdangerous permissions and collecting suspicious data from them, then we have begun tocategorize them. And after thoroughly examining them and identifying a few batches of databotnets, they became supervised to implement machine learning algorithms on them as labels.At the data collection stage, all categories of Google Play data were extracted. Finally, due tothe different number of labels(not balance) in each data set, and in order to compare the workwith previous tasks, the accuracy of their work and then the kappa criterion for work evaluationwere considered and the results reported.

android malware_ machine learning_ botnet detection_ malware detection