Abstract. We introduce StPowla, a workflow based approach to business process modelling that integrates a simple graphical notation, to ease the presentation of the core business process, a natural policy language, Appel, to provide the necessary adaptation to the varied expectations of the various business stakeholders, and the Service Oriented Architecture, to assemble and orchestrate available services in the business process. We illustrate the approach with a loan approval process.

Bike-sharing systems are a means of smart transportation in urban environments with the benefit of a positive impact on urban mobility. In this paper we are interested in studying and modeling the behavior of features that permit the end user to access, with her/his web browser, the status of the Bike-Sharing system. In particular, we address features able to make a prediction on the system state. We propose to use a machine learning approach to analyze usage patterns and learn computational models of such features from logs of system usage.\ud On the one hand, machine learning methodologies provide a powerful and general means to implement a wide choice of predictive features. On the other hand, trained machine learning models are provided with a measure of predictive performance that can be used as a metric to assess the cost-performance trade-off of the feature. This provides a principled way to assess the runtime behavior of different components before putting them into operation

Bike-sharing\ud systems\ud (BSS)\ud are\ud a\ud means\ud of\ud smart\ud transportation\ud with\ud the\ud benefit\ud of\ud a\ud positive\ud impact\ud on\ud urban\ud mobility.\ud To\ud improve\ud the\ud satisfaction\ud of\ud a\ud user\ud of\ud a\ud BSS,\ud it\ud is\ud useful\ud to\ud inform\ud her/him\ud on\ud the\ud status\ud of\ud the\ud stations\ud at\ud run\ud time,\ud and\ud indeed\ud most\ud of\ud the\ud current\ud systems\ud provide\ud the\ud information\ud in\ud terms\ud of\ud number\ud of\ud bicycles\ud parked\ud in\ud each\ud docking\ud stations\ud by\ud means\ud of\ud services\ud available\ud via\ud web.\ud However,\ud when\ud the\ud departure\ud station\ud is\ud empty,\ud the\ud user\ud could\ud also\ud be\ud happy\ud to\ud know\ud how\ud the\ud situation\ud will\ud evolve\ud and,\ud in\ud particular,\ud if\ud a\ud bike\ud is\ud going\ud to\ud arrive\ud (and\ud vice\ud versa when\ud the\ud arrival\ud station\ud is\ud full).\ud To\ud fulfill\ud this\ud expectation,\ud we\ud envisage\ud services\ud able\ud to\ud make\ud a\ud prediction\ud and\ud infer\ud if\ud there\ud is\ud in\ud use\ud a\ud bike\ud that\ud could\ud be,\ud with\ud high\ud probability,\ud returned\ud at\ud the\ud station\ud where\ud she/he\ud is\ud waiting.\ud The\ud goal\ud of\ud this\ud paper\ud is\ud hence\ud to\ud analyze the\ud feasibility\ud of\ud these\ud services.\ud To\ud this\ud end,\ud we\ud put\ud forward\ud the\ud idea\ud of\ud using\ud Machine\ud Learning\ud methodologies,\ud proposing\ud and\ud comparing\ud different\ud solutions

Abstract. Appel is a general language for expressing policies in a variety of application domains with a clear separation between the core language and its specialisation for concrete domains. Policies can conflict, thus leading to undesired behaviour. We present a novel formal semantics for the Appel language based on ∆DSTL(x) (so far Appel only had an informal semantics). ∆DSTL(x) is an extension of temporal logic to deal with global applications: it includes modalities to localize properties to system components, an operator to deal with events, and temporal modalitiesà la Unity. A further contribution of the paper is the development of techniques based on the semantics to reason about conflicts.