Malware prediction algorithm: Systematic review

• Main Authors: Mahrin, Mohd. Naz’Ri, Chuprat, Suriayati, Subbarao, Anusuyah, Mohd. Ariffin, Aswami Fadillah, Talib, Mohd. Zabri Adil, Ahmad Darus, Mohammad Zaharudin, Abd. Aziz, Fakhrul Afiq
• Format: Article
• Published: Little Lion Scientific 2018
• Subjects: T Technology (General)
• Online Access: http://eprints.utm.my/id/eprint/84467/; http://www.jatit.org/volumes/Vol96No14/28Vol96No14.pdf
• Institution: Universiti Teknologi Malaysia

Malware is a threat to information security and poses a security threat to harm networks or computers. Not only the effects of malware can generate damage to systems, they can also destroy a country when for example, its defense system is affected by malware. Even though many tools and methods exist, breaches and compromises are in the news almost daily, showing that the current state-of-the-art can be improved. Hundreds of unique malware samples are collected on a daily basis. Currently, the available information on malware detection is ubiquitous. Much of this information describes the tools and techniques applied in the analysis and reporting the results of malware detection but not much in the prediction on the malware development activities. However, in combating malware, the prediction on malware behavior or development is as crucial as the removing of malware itself. This is because the prediction on malware provides information about the rate of development of malicious programs in which it will give the system administrators prior knowledge on the vulnerabilities of their system or network and help them to determine the types of malicious programs that are most likely to taint their system or network. Thus, based on these, it is imperative that the techniques on the prediction of malware activities be studied and the strengths and limitations are understood. For that reason, a systematic review (SR) was employed by a search in 5 databases and 89 articles on malware prediction were finally included. These 89 articles on malware prediction has been reviewed, and then classified by techniques proposed in detection of new malware, the identified potential threats, tools used for malware prediction, and malware datasets used. Consequently, the findings from the systematic review can serve as the basis for a malware prediction algorithm in future as malware predication became a critical topic in computer security.