Despite the continued advancements in security research, malware persists as being a major threat in this digital age. Malware detection is a primary defence strategy for most networks but the identification of malware strains is becoming increasingly difficult. Reliable identification is based upon characteristic features being detectable within an object. However, the limitations and expense of current malware feature extraction methods is significantly hindering this process. In this paper, we present a new method for identifying malware based on behavioural feature extraction. Our proposed method has been evaluated using seven classification methods whilst analysing 2,068 malware samples from eight different families. The results achieved thus far have demonstrated promising improvements over existing approaches.