Web service composition can be adopted to\ud develop information systems through integration of services\ud to obtain complex composed services. While interfaces of\ud services are known at composition time, the quality of a\ud composed service may depend on the ability of its\ud component services to react to unforeseen situations, such\ud as data quality problems and service coordination problems.\ud In this work, we propose an approach to analyze the quality\ud of composed services using fault injection techniques, by\ud inspecting the reaction of a composed process to injected\ud faults; the aim is to assess the process quality in terms of\ud fault monitoring and, more generally, fault tolerance\ud capabilities. The component services are analyzed either\ud as black-boxes, when only input and output messages are\ud considered or as white-boxes, when data sources used by\ud services are considered. A test bed is illustrated on a\ud selected example, and results of extensive testing are\ud discussed and framed into a process analysis methodology

No abstract

Abstract. During Web service discovery phase, the policy selection step must occur to retrieve not only the Web services able to perform the required functionalities, but also Web services able to ensure a given quality level. Both service providers and service users define policies. The former list the supported service features and use a technical vocabulary (e.g., bandwidth, latency) whereas the latter identify the requirements using high level vocabulary closer to human users (e.g., speed). The aim of this paper is twofold. On one hand, it introduces a model for expressing quality according to both applications and human users perspectives. Such a model, compliant with the WS-Policy framework, not only mediates between the application and human user perspectives, but is also capable of considering the different importance assigned by the user to quality dimensions. On the other hand, the paper introduces a policy selection model based on the adopted quality model. According to this approach, human users can express quality requirements according to a high level language; these requirements are then matched against low level specifications of the service quality.

No abstract

Accidents in work areas can be prevented using risk management systems able to identify the risk events which usually precede accidents, and able to enact prevention mechanisms to avoid them whenever possible. Most accidents are announced by risk events, which may be identified by monitoring the environment and managed through preventive strategies. Otherwise, an accident determines an emergency situation that should be addressed through corrective strategies. In this paper, we propose a service-based solution for a Risk Management System (RMS) exploiting the explicit definition and computation of risks, based on the introduction of revealing and protection devices. Furthermore, motivating examples of risk events are introduced to illustrate the fundamentals of our prototype system.