Temporal Service Level Agreements (SLAs) are a distinguishing characteristics for offered (composed) services: the weaker the restrictions, the broader the ranges of input parameters, the wider the applicability of an offered service. We propose a technique for the providers of composed services with temporal parameters to compute the broadest ranges for the temporal input parameters such that all service invocations where the parameters are within these ranges are dynamically controllable, i.e. that the adherence to temporal constraints can be guaranteed.

Inter-organizational business processes can be designed top-down: starting from a global process model, where each activity and gateway is assigned to one of the participating organizations, local processes (or views/stubs for refinement) are generated for each organization. The resulting inter-organizational process is executed by the local processes in a fully distributed manner. Implementing the data-flow in such a scenario is highly complex, both because of the combinatorial explosion of possible solutions and because of conflicting goals for implementing the data-flow. We present a flexible, heuristic algorithm for implementing the data-flow for inter-organizational processes according to various user preferences, which significantly improves over the base-line approaches proposed so far. The approach is fully implemented and a comprehensive, multi-dimensional evaluation is provided.

The composition of web services frequently requires the adequate representation of temporal requirements, temporal obligations, and temporal assumptions. Currently, the modeling of orchestrations and their service level agreements including temporal constraints suffers however from a mismatch between more procedural orchestration models and more declarative formulations of temporal constraints. We proposed the introduction of temporal conditions in the formulation of conditional constructs for process modeling to give process designers explicit control over the temporal behavior of the processes they model. By transforming the so modeled processes to processes with decision nodes we can take advantage of the recently developed CSTNUDs, temporal control networks with conditions, uncertainty and decisions, which in particular provide for checking the dynamic controllability of temporally constrained processes.

Inter-organizational business processes can be designed top-down following the public to private approach. First, a global process model is designed. In a second step, each activity is assigned to one of the business partners. Finally, processes (or views/stubs for refinement) are generated for each partner. The resulting inter-organizational process is executed in a fully distributed way by the set of communicating local processes. Existing approaches concentrate on the control-flow and ignore the data-flow or follow rather simple approaches for implementing the data-flow. However, automatically generating message-exchanges is a challenging task since numerous solutions with hugely varying advantages and disadvantages are possible. In this paper, we present criteria for assessing data-flow implementation and discuss baseline approaches that each optimize one of these criteria. All algorithms are implemented and evaluated for a variety of different process classes.

Interorganizational business processes aim to integrate local processes to support the seamless cooperation of organizations. Process view approaches are an adequate method balancing the communication requirements for enabling collaboration and the required privacy hiding internals of private business processes. We focus on a top-down scenario for the development of interorganizational processes, where first an abstract global process is designed. Then each step of the global process is assigned to one of the partners and a local process is generated for each partner as a view on the global process. Finally, the partners implement or adopt their processes based on their views. We present an algorithm for the fully automatic generation of views for any block-structured input process with arbitrary partner assignments, provide a method for merging the partner's views to reconstruct the global process and prove the correctness of the view generation method.

Many recursive query processing applications are still poorly supported, partly because implementations of general recursive capabilities are inefficient and hard to understand for users, partly because the approaches do not integrate well with existing query languages. An extension is proposed of the database language SQL for the processing of recursive structures. The new constructs are integrated in the view definition mechanism of SQL. Therefore, users with knowledge of SQL can take advantage of the increased functionally without learning a new language. The construct is based on a generalization of transitive closure and is formally defined. Because of the importance of extreme value sections, special constructs are introduced for the selection of tuples with minimal or maximal values in some attributes. Applying these selections on recursively defined views constitutes nonlinear recursion. By the introduction of special constructs for these selections, dealing with general nonlinear recursion can be avoided.>