Sometimes, explicit relationships between entities do not provide sufficient information or can be unavailable in the real world. Unseen latent relationships may be more informative than explicit relationships. Thereby, we provide a method for constructing latent informative links between entities, using their common features, where entities are regarded as vertices on a graph. First, we employ a hierarchical nonparametric model to infer shared latent features for entities. Then, we define a filter function based on information theory to extract significant features and control the density of links. Finally, a couple of stochastic interaction processes are introduced to simulate dynamics on the net-works so that link strength can be retrieved from statistics in a natural way. In experiments, we evaluate the usage of filter function. The results of two examples based on mixture networks show how our method is capable of providing latent informative relationships in comparison to explicit relationships
IOS Press, Inc.
6751 Tepper Drive
Clifton, VA 20124
Tel.: +1 703 830 6300
Fax: +1 703 830 2300 firstname.lastname@example.org
(Corporate matters and books only) IOS Press c/o Accucoms US, Inc.
For North America Sales and Customer Service
West Point Commons
Lansdale PA 19446
Tel.: +1 866 855 8967
Fax: +1 215 660 5042 email@example.com