A Fuzzy Matching with Reasoning Publish/Subscribe System based on Ontology

Traditional publish/subscribe systems suffer from either the limited expressiveness of events or the complexity of matching algorithms. Moreover, most of them do not consider the semantic information of events. This study proposes a fuzzy matching with reasoning publish/subscribe system, which employs ontology to model events and subscriptions. The study elaborate the two main components in the system: 1) Offline part in which domain ontology defines the concepts of events and the relationships between them, and extra knowledge is a supplement for the domain ontology. 2) Online part which shows the details of published events, subscription requests and matching between published events and subscription requests, including how to translate events to RDF event graph, how to translate subscription requests to SPARQL query, and details of the matching algorithm. This study combines traditional publish/subscribe system with semantic technology, and thereby improves the capability of expressiveness and accuracy of matching algorithm.