Design of an autonomous framework for efficient large scale management of next generation web service mashups

Abstract

The objective of this PhD research is the study and development of an autonomous framework focusing on the dynamic composition and deployment of service mashups. By enriching the available building blocks with semantic descriptions, automatic composition of new service mashups is achieved through the use of planning algorithms. The composed mashups will be automatically deployed on the available resources making optimal use of bandwidth, storage and computing power of the network and server elements. This system will also be extended with dynamic recovery from external errors, such as network failure or services delivering erroneous results. The last step will encompass the automatic extraction of new knowledge from request patterns, behavior and feedback of users for the optimization of future requests. A number of relevant case studies will be defined in the domain of eHealth and Web 3.0, since the use of workflows for efficient data processing in these areas is very important. © 2009 IFIP International Federation for Information Processing.