Self-organized Detection of Relationships in a Network
Multistate operations within a network result in high-dimensional, multivariate temporal data, and are useful for systems, which monitor access to network entities like resources, objects, etc. Efficient self organization of such multistate network operations stored in databases with respect to relationships amongst users or between a user and a data object are an important and a challenging problem. In this work, a layer is proposed where discovered relationship patterns amongst users are classified as clusters. This information along with attributes of involved users is used to monitor and extract existing and growing relationships.