Abstract: The proliferation of Android applications has revolutionized the way we interact with mobile technology, offering unparalleled convenience and functionality. However, this rapid expansion has also given rise to a pressing concern: the proliferation of Android malware. Malicious actors exploit the open nature of the Android platform to distribute harmful applications, posing significant threats to users' privacy, security, and data integrity. Current scenarios reveal a multitude of tactics employed by malware authors, including disguised applications, phishing scams, and data exfiltration techniques, exacerbating the complexity of malware detection and classification. In response to these challenges, this study proposes a novel approach for Android malware detection and classification leveraging Support Vector Machines (SVM) and the K-means algorithm. The methodology encompasses several critical stages: application scan, application list extraction, feature extraction, and access information extraction. Through these processes, comprehensive data is collected and analyzed to discern patterns indicative of malicious behavior. SVM, renowned for its effectiveness in supervised learning tasks, is employed to classify applications based on extracted features, while K-means clustering facilitates unsupervised classification, augmenting the detection capabilities of the system.

Experimental evaluation on a diverse dataset underscores the efficacy of the proposed methodology in accurately identifying and classifying Android malware applications. Our results showcase impressive performance metrics, including high accuracy, precision, recall, and F1-score, affirming the robustness of the approach in the face of evolving malware threats. By addressing the current challenges in Android malware detection and classification, this research contributes to the advancement of cybersecurity measures in the mobile ecosystem. Looking ahead, further research is warranted to enhance the scalability and adaptability of the proposed approach to evolving malware landscapes. Additionally, collaboration among researchers, industry stakeholders, and policymakers is essential to foster a proactive and collaborative approach to combating Android malware and safeguarding user privacy and security in the digital age.

Keywords: Android malware, SVM (Support Vector Machines), K-means algorithm, Malicious applications, Mobile security, Data privacy, Data integrity, Malware detection, Classification, Supervised learning, Unsupervised learning, Feature extraction, Experimental evaluation, Performance metrics, Cybersecurity measures etc.

| DOI: 10.17148/IARJSET.2024.11452