M2VMapper: Malware-to-Vulnerability mapping for Android using text processing
作者：
Highlights：
• Automatic mapping of malware to exploited vulnerabilities in Android.
• Considered 150 malware and 9 different types of vulnerabilities affecting Android.
• Leveraged pre-trained learning models, where the word embeddings are the key for feature engineering.
• Achieved 99.81% accuracy using Deep Learning models.
• First-of-its kind study that can aid enhancement of Android security in development phase.
摘要
•Automatic mapping of malware to exploited vulnerabilities in Android.•Considered 150 malware and 9 different types of vulnerabilities affecting Android.•Leveraged pre-trained learning models, where the word embeddings are the key for feature engineering.•Achieved 99.81% accuracy using Deep Learning models.•First-of-its kind study that can aid enhancement of Android security in development phase.
论文关键词：Android,Deep Learning,Malware,Mapping,Language model,Vulnerability
论文评审过程：Received 9 November 2020, Revised 18 August 2021, Accepted 29 November 2021, Available online 7 December 2021, Version of Record 10 December 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.116360
