Malware Dataset Generation and Evaluation

With the rapid growth of technology and IT-enabled services, the potential damage caused by malware is increasing rapidly. A large number of detection methods have been proposed to arrest the growth of malware attacks. The performance of these detection methods is usually established using raw or feature datasets. The non-availability of adequate datasets often becomes a bottleneck in malware research. To address this issue, this paper presents two malware feature datasets on two different platforms to support validation of the effectiveness of a malware detection method. We evaluate the usefulness of our datasets in a supervised framework.