MALWARE ANALYSIS AND DETECTION FOR MICROSOFT TECHNOLOGIES

Abstract

Malware detection is always a hot issue and a priority task in cyber crimes. Despite a lot of work in the past malware detection in Microsoft Word is being a major challenge for researchers and other practitioners. This research examines the malware and detects the malicious files with the help of structural path features and lexical based features of extracted URL from unzipped XML files of Microsoft Word. This research carried out three experiments and finally reached to a goal with 0.97% accuracy with a highest true positive rate of 0.98% and lowest false positive rate of 0.012%. It showed a somehow reduced TPR rate in detecting benign files but can be increase in future while doing more precise work upon malicious URL used in documents.