An algorithm with PN to improve recall of service matching

Traditional UDDI does not have the capability to process semantic and match to the service attribute, which results in a low recall rate and matching rate. Through the service matching framework of Multi-agent, a Petri net based OWL-S semantic matching mechanism is proposed to de al with the shortcoming, in which PNSDL description language was used to publish and request services. In virtue of the PN´s ability for describing we b service, the matching mechanism uses the possibility and necessity to quantify the confidence levels that the service could satisfy a request, so as to finish the matching of service. The matching mechanism has the advantage of powerful ability to process, model for Multiservice, high reality of matching and high matching rate, etc.