DETECTION OF ANDROID BOTNETS
- Version
- Download 62
- File Size 817.44 KB
- File Count 1
- Create Date 1 June 2024
- Last Updated 1 June 2024
DETECTION OF ANDROID BOTNETS
Dr. Deepak A. Vidhate 1
1Professor and Head, Information Technology, Dr. Vithalrao Vikhe Patil College of Engineering, Ahmednagar
Prof. A.A.Pund 2
2Assistant Prof, Information Technology, Dr. Vithalrao Vikhe Patil College of Engineering, Ahmednagar
Harshvardhan Avachar 3, Yashanjali Berad 4, Shubham Bhapkar 5, Sakshi Parkhe 6
3,4,5,6 Department of Information Technology, Dr. Vithalrao Vikhe Patil College of Engineering, Ahmednagar
Abstract - In the dynamic landscape of cybersecurity, the persistent proliferation of botnets remains a formidable challenge. Conventional detection methods often grapple with elevated false positive rates and struggle to keep pace with the evolving tactics employed by malicious actors. This study delves into the domain of botnet detection, with a primary focus on refining the identification process for HTTP-based botnets. Our approach harnesses the power of K-Nearest Neighbors (KNN) and Logistic Regression algorithms, strategically amalgamating their capabilities to navigate the intricate digital terrain. Departing from prior studies, we directly confront the botnet detection challenge by synergizing two robust machine learning techniques. Through the fusion of KNN and Logistic Regression, our system achieves unparalleled accuracy and efficiency in discerning HTTP botnet activities. Furthermore, this research pioneers a groundbreaking methodology for botnet detection, centering around the innovative fusion of TFIDF and Textrank algorithms. This hybrid approach significantly bolsters the precision of information extraction and summarization. In an age inundated with vast digital datasets, our method adeptly sifts through information while furnishing concise and relevant summaries, thereby conserving invaluable time and resources. The distinctive aspect of our approach lies in its ability to swiftly adapt to emerging botnet strategies. Through extensive testing and comparative analysis against existing models, our system surpasses prior methodologies, demonstrating its proficiency in accurately identifying botnet activities while minimizing false positives. Furthermore, our solution serves as an exemplar of efficiency, facilitating prompt and effective identification of pernicious botnets that imperil data security
Key Words: Botnet Detection, Cybersecurity, K-Nearest Neighbors (KNN), Logistic Regression, TFIDF, Textrank, Information Summarization, Digital Security, Malware Detection.