ISSN 0021-3454 (print version)
ISSN 2500-0381 (online version)
Menu

5
Issue
vol 60 / MAY, 2017
Article
UDC 004.056

METHOD OF STRUCTURAL FEATURE EXTRACTION FOR MALWARE DETECTION

D. V. Komashinckiy
St. Petersburg Institute for Informatics and Automation of Russian Academy of Sciences, Laboratory of Computer Security Problems; Post-Graduate Student


I. V. Kotenko
St. Petersburg Institute for Informatics and Automation of Russian Academy of Sciences, Laboratory of Computer Security Problems ; Professor


Abstract.   The approach to malicious documents detection based on the application of machine learning, previously proposed by the authors, is applied to development of a method for extraction of static structural features from a data stream. The method makes use of structural features of malicious document elements and linkage between them.
Keywords: feature extraction, malware, classification.