HEFESTDROID: Highly Effective Features for Android Malware Detection and Analysis

Rapid globalization and advances in mobile technology have brought about phenomenal attention great opportunities for android application developers to contribute meaningfully the global digital market. The platform being one of famous operating systems has highest number applications market with a total share 76.23% between August 2018 2019, according report stats counter. However, substantial on led malware attacks user’s privacy sensitive documents. Consequently, significant detection studies been carried out reduce attacks. This paper analyses impact using highly effective permission features decipher problem attack. Highly Effective Features Android Malware Detection Analysis (HEFEST) summarises four be considered conducting analysis classifications. recognized this study are; Normal Declared Permission, Dangerous Signature-Based Signature-or-system. selection is based capabilities depicting behaviors apps. research data are drawn from Drebin open source, dataset comprises 15,036 benign malicious extracted 215 distinct features, records 9,026 were 6,010 applications. compares accuracy machine learning-based algorithms; Support Vector Machine, K-Nearest Neighbor achieve comprehensive ratio detection, classifier strong decision classification exceptional computational efficiency. model correctly classified 2,812 2,869 appropriately an 98.0% also 1,607 1,642 accurately success rate 97.9%. Generally, was archived.