A comprehensive survey on hardware-assisted malware analysis and primitive techniques

A Novel Approach of Evasive Malware Analysis Through Binary Opcode and BERT

“…Recent research in malware analysis has shifted towards more innovative approaches, leveraging the advancements in machine learning and artificial intelligence. Among these, the application of binary opcode analysis has gained traction, offering a more granular view of the malware's operational code and behavior [11,12]. However, the complexity and volume of opcode sequences in modern malware require approaches that can effectively interpret and classify vast amounts of data with high precision [11].…”

“…Among these, the application of binary opcode analysis has gained traction, offering a more granular view of the malware's operational code and behavior [11,12]. However, the complexity and volume of opcode sequences in modern malware require approaches that can effectively interpret and classify vast amounts of data with high precision [11]. The introduction of BERT (Bidirectional Encoder Representations from Transformers) in natural language processing has opened new avenues for research, yet its application in the realm of malware analysis is still in nascent stages [13].…”

“…The introduction of BERT (Bidirectional Encoder Representations from Transformers) in natural language processing has opened new avenues for research, yet its application in the realm of malware analysis is still in nascent stages [13]. Current research has shown promise but also reveals gaps, particularly in the context of evasive malware detection, where the adaptability and evolving nature of threats pose unique challenges [11,2,3]. These inadequacies in current research highlight the need for a more robust and adaptive approach that can evolve alongside the malware it seeks to analyze.…”