Semantic-driven model of dynamic logistic services composition and optimization based on QoS

In mass services environment, how to schedule resources efficiently and reliably in the Fourth Party Logistics is a hot topic recently. Semantic web and web service technology are introduced into our project on the background of 4PL. The starting point of our work is to define service conceptions, semantic matching and reasoning in logistic. Then, two-stage solution algorithm (TSSA) is proposed which always can figure out solutions with low time-consuming and makes the platform more interactive. In the 1^st stage, a semantic-driven dynamic logistic services composition algorithm is offered in the project, which is the improvement of classical Ant Colony algorithm. It can figure out a couple of logistic service paths at one time and the paths are the input of the 2^nd stage. TSSA optimize the results based on Q OS and complexity of TSSA is polynomial-time which is much better than that of classical Ant Colony algorithm.. In order to improve calculation for QoS, a trust and reputation enhanced QoS calculation method is presented on the light of Bayes learning theory. At last, with implementing TSSA in different logistic environments and compared its results with that of classical Ant Colony algorithm, the results show that TSSA can quickly and effectively get solutions meeting customers´ requirements well.