Graph-Based Publish/Subscribe Research with Dynamic Subscriptions

The publish/subscribe paradigm is well-regarded for its ability to successfully disseminate data while decoupling data providers from data consumers. Traditionally, pub/sub uses two common forms of filtering: topic-based and content-based, with a matching language that is limited to those filtering forms. However, it can be difficult to use these approaches to express interest in a semantic information domain. The current, past, and future research work shown in this paper explores an alternate publication and filtering model based on graphs, where consumers can express their interest in certain publications by subscribing to parts of a graph. This graph-based approach takes advantage of the properties of nodes and semantics of edges to model an application domain. In our research we have developed a graph-based pub/sub middleware prototype and are currently working on extending it to support dynamic graph updates and dynamic subscriptions.