A Survey on Malware Detection based on API Calls
Kaixin Chang, Nailiang Zhao, Liang Kou · 2022 9th International Conference on Dependable Systems and Their Applications (DSA) · 2022
Malware has posed serious security threats to individuals, public systems, and corporations worldwide. Therefore, how to detect malware early, preferably block it beforehand, has always been a concern for the industry and academia. Previous studies have shown that executed API calls can represent malicious behaviors. Methods leveraging both API call information and machine learning or deep learning models are proved effective and have relatively higher detection rates, thus have quickly become the hotspot of research. This survey introduces recent works about malware detection based on API calls, including the overall process of malware detection tasks, relevant state-of-the-art works using machine learning and deep learning models, and four challenges in malware detection based on API calls. It also provides some suggestions for future enhancement of malware detection methods.