Malicious software (malware) attacks are on the rise with the explosion of Internet of Things (IoT) worldwide. With the proliferation of Big Data, it becomes a time consuming process to use various automatic approaches and techniques that are available to detect and capture malware thoroughly. Visualisation techniques can support the malware analysis process for performing the similarity comparisons and summarisation of possible malware in such Big Data contexts. In this paper, we design a novel classification of malware using visualization of similarity matrices. The prime motivation of our proposal is to detect unknown malwares that undergo the innumerable obfuscations of extended x86 IA-32 (opcodes) in order to evade from traditional detection methods. Overall, the high accuracy of classification achieved with our proposed model can be observed visually due to significant dissimilarity of the behaviour patterns exhibited by malware opcodes as compared to benign opcodes.