The goal model at the core of the goal-oriented approach to requirements engineering graphically represents relationships between the goals (functional requirements), tasks (realizations of goals), and softgoals (nonfunctional properties) involved in designing a system. It may, however, be impossible to find a design that fulfills all top-level goals and satisfies all softgoals. In such cases, it is useful to find designs that provide the required functionality while satisfying the most preferred set of softgoals under the goal model's constraints. Existing methods typically consider quantitative preferences over softgoals, where the quantification produces a ranking among softgoals. We instead present an approach that considers expressive qualitative preferences over softgoals; unlike quantitative preferences, these can model interacting and/or mutually exclusive subgoals. Our framework employs a model checking-based method for reasoning with qualitative preferences to identify the most preferred alternative(s). We evaluate our approach using existing goal models from the literature. AbstractThe goal model at the core of the goal-oriented approach to requirements engineering graphically represents relationships between the goals (functional requirements), tasks (realizations of goals), and softgoals (non-functional properties) involved in designing a system. It may, however, be impossible to find a design that fulfills all top-level goals and satisfies all softgoals. In such cases, it is useful to find designs that provide the required functionality while satisfying the most preferred set of softgoals under the goal model's constraints. Existing methods typically consider quantitative preferences over softgoals, where the quantification produces a ranking among softgoals. We instead present an approach that considers expressive qualitative preferences over softgoals; unlike quantitative preferences, these can model interacting and/or mutually exclusive subgoals. Our framework employs a model checkingbased method for reasoning with qualitative preferences to identify the most preferred alternative(s). We evaluate our approach using existing goal models from the literature.

No abstract

Abstract. In most client-server interactions over the Web, the server requires the client to disclose certain credentials before providing the client with the requested service (server policy). The client, on the other hand, wants to minimize the sensitivity of the set of credentials disclosed (client preference). We present a qualitative preference formalism based on conditional importance networks (CI-nets) for representing and reasoning with client preferences over the relative sensitivity of sets of credentials. The semantics of CI-net preferences is described using a preference graph over the set of credentials for which the preferences are expressed. We develop a model checking-based approach for analyzing the preference graph, efficiently verifying whether one set of credentials is more sensitive than another (dominance testing). Further, we identify the least (minimum) sensitive set of information that may be disclosed by the client to get access to the desired service. We present a technique based on iterative verification and refinement of the preference graph for computing a sequence of credential sets, ensuring that a credential set with higher sensitivity is never returned before one with lower sensitivity. We present a prototype implementation and preliminary simulation results.

Abstract. In many practical applications, trade-offs involving non-functional attributes e.g., availability, performance play an important role in selecting component services in assembling a feasible composition, i.e., a composite service that achieves the desired functionality. We present TCP-Compose , an algorithm for service composition that identifies, from a set of candidate solutions that achieve the desired functionality, a set of composite services that are non-dominated by any other candidate with respect to the user-specified qualitative preferences over non-functional attributes. We use TCP-net, a graphical modeling paradigm for representing and reasoning with qualitative preferences and importance. We propose a heuristic for estimating the preference ordering over the different choices at each stage in the composition to improve the efficiency of TCP-Compose . We establish the conditions under which TCP-Compose is guaranteed to generate a set of composite services that (a) achieve the desired functionality and (b) constitute a non-dominated set of solutions with respect to the user-specified preferences and tradeoffs over the nonfunctional attributes.

Many approaches to the Web service composition problem benefit from their use of formal methods to guarantee the correctness of the composite services that they produce, but these approaches often require the functionality of the composite service to be specified using one particular formalism (e.g., goal graphs, temporal logic, pre-/post-conditions). As a result, each of these existing approaches falls short in realizing a composite service when the required functionality cannot be fully expressed in the supported formalism. To overcome this problem, we propose a new formal meta-framework that (a) is capable of reusing any existing formalisms and (b) allows the use of functional requirements that are currently not expressible in any one formalism. Our technique assumes that any functional requirement can be decomposed and expressed as a boolean combination of "atomic" requirements, which are representable in at least one existing formalism. Based on this assumption, we use existing methods to identify sets of Web services that conform to the atomic requirements. Our meta-framework then identifies compositions that conform to the overall functional requirement by (a) employing satisfiability techniques to prune the (exponentially large) space of possible compositions and (b) building workable compositions from satisfiable sets of services. As a result, our meta-framework allows for easy and effective memorization of prior composition results, thereby enhancing the efficiency of generating new compositions.