Novel malware detection methods by using LCS and LCSS
Fahad Mira, Antony Brown, Wei Huang · 2016
The field of computer security faces numerous vulnerabilities which cause network resources to become unavailable and violate systems confidentiality and integrity. Malicious software (Malware) has become one of the most serious security threats on the Internet. Malware is a widespread problem and despite the common use of anti-virus software, the diversity of malware is still increasing. A major challenge facing the anti-virus industry is how to effectively detect thousands of malware samples that are received every day. In this paper, a novel approach based on dynamic analysis of malware is proposed whereby Longest Common Subsequence (LCSS) and Longest Common Substring (LCS) algorithms are adopted to accurately detect malware. The empirical results show that the proposed approach performs favorably compared to other related work that use API call sequences.