A bayesian belief network for data leakage investigation

Today, it is common to share and to disseminate digital data in both the commercial and personal domains. Incidents of data leakage in which personal or commercial sensitive data is leaked to untrusted or non-intended recipients are also on the rise. Although researches on data leakage detection are emerging, there is little research on the detection of data leakage from the perspective of evidence analyses. On the bases of the probabilities analyses of evidence, this paper aims to use a Bayesian Belief Network (BBN) to detect the likelihood of the data leakage path and the roles of corresponding parties. Through the application of the BBN to a notorious criminal case in Hong Kong, the "Sexy Photos Gate", which involved the leakage of intimate digital photographs of celebrities in Hong Kong, this paper will illustrate the most possible leakage path and the responsible parties for the leakage.