Efficient matching for web-based publish/subscribe systems

Abstract

There is a need for systems being able to capture the dynamic aspect of the web information by notifying users of interesting events. Content-based publish/subscribe systems axe an emerging type of publish/subscribe systems where events axe filtered according to their attribute values, using filtering criteria defined by the subscribers, and then sent to the interested subscribers. Compaxed to traditional publish/subscribe systems, content-based systems offer more subscription expressiveness. The cost of this gain in expressiveness is an increase in the complexity of the matching process: the more sophisticated the constructs, the more complex the matching process. In this paper, we present an efficient and scalable solution to the matching problem. We also present a semi-structured event model which is well suited for the information published on the Web, and flexible enough to support easy integration of publishers.