Techniques are disclosed relating to data management. A computer system may evaluate network traffic to extract and group data objects based on their content satisfying similarity criteria, and to identify baseline behavior with respect to those data objects. The computer system may generate data-defined network (DDN) data structures that include a content class and one or more behavioral classes. The content class may be indicative of one or more of the data objects that have been grouped based on them satisfying the similarity criteria. The one or more behavioral classes may indicate baseline behavior of those data objects within the content class as determined from evaluation of the network traffic. The computer system may detect, using the DDN data structures, anomalous data behavior within network traffic. In response to detecting anomalous data behavior, the computer system may prevent network traffic corresponding to the anomalous data behavior from being communicated.