Section: New Results

Expressiveness of computational models

Participants : Roberto Amadini, Ornela Dardha, Maurizio Gabbrielli, Daniel Hirschkoff, Jean-Marie Madiot, Jacopo Mauro, Davide Sangiorgi, Gianluigi Zavattaro.

Expressiveness refers to the study of the descriptive power of computational models.

The fusion calculi are a simplification of the $\pi$-calculus in which input and output are symmetric and restriction is the only binder. We show [35] a major difference between these calculi and the $\pi$-calculus from the point of view of types, proving some impossibility results for subtyping in fusion calculi. We propose a modification of fusion calculi in which the name equivalences produced by fusions are replaced by name preorders, so to be able to import subtype systems, and related results, from the $\pi$-calculus. We have studied the consequences of the modification on behavioural equivalence and expressiveness.

In Focus we use notions of constraint to define in a succinct way models of computation and current constraint solving technologies to solve problems modeled using constraints. For this reason we have studied the expressive power of various computational models involving constraints and their practical impact in terms of solving/execution performances. In [18] we have investigated how the notion of constraint augments the expressive power of a concurrent language if priorities are introduced. The chosen language is Constraint Handling Rules, a committed-choice declarative language originally designed for writing constraint solvers and that is nowadays a general purpose language. The result has been otbained by first formalising the meaning of language encodings and language embedding, widely used in concurrency theory. Different ways to model and define disaster scenarios are analyzed and compared in [11] , where we study a model expressive enough to define a disaster scenario that, at the same time, can be used to find plans to save the victims of a disaster using modern constraint solving technology. Similarly, different computation models are considered in [22] where we study how machine learning techniques can be used to boost the performances of constraint solvers. A technique dubbed “portfolio approach” is used to combine the different performances of constraint solvers to obtain a globally better solver using, as a starting point, a simple low-level constraint language.

In [30] we propose an integration of structural sub-typing with boolean connectives and semantic sub-typing to define a Java-like programming language that exploits the benefits of both techniques. The resulting language has a more expressive set of types that comes from the use of boolean constructs, negation types, and the integration of structural and nominal sub-typing in an object-oriented setting. By implementing traditional Java-language constructs we show that the proposed language is also expressive enough w.r.t. the Java language.