User-Transition Behavior Analysis in Data Publication Using Bipartite Graph

Users requesting trusted third parties to publish their sets of data items on transitions to access different services have become very popular in this information technology era. We therefore need to understand the trends or behavior of the users in making transitions to states for services, while publishing their set of non-sensitive data items. First and foremost, we constructed a user-state weighted bipartite graph model which is not vulnerable to attacks by adversaries. Using this model, the user-transition behavior were analyzed mainly from four viewpoints; user interest choice, user interest concentration, user transition attributes and user-transition importance. Two data sets which contain two 30-minutes time intervals were used; from 16:00:00 to 16:29:59 and from 16:30:00 to 16:59:59. In order to study the user-transition characteristics for the different services at the states, the states were classified into thirteen types based on the data sets. Finally, we illustrated the attributes of state's access behavior and revealed the variations of user-transition behavior.