First International Workshop on

Component-based Software Development 
in Computational Logic (COCL)

Saturday, September 19, 1998 - Pisa (Italy)

Abstracts of the talks

Composing complete and partial knowledge
Sofie Verbaeten (K.U.Leuven, Belgium)
Annalisa Bossi (Universita' Ca' Foscari di Venezia, Italy)
The representation of knowledge in the logic OLP-FOL is split in two parts: writing definitions for known concepts, and writing constraints, expressing partial knowledge on other concepts. This is reflected in an OLP-FOL theory T, which is a pair: T = (T_d, T_c). The definition part T_d contains the definitions for known predicates in the form of a normal open logic program (OLP), whereas the first order logic (FOL) part T_c is a set of FOL axioms, expressing partial knowledge on other predicates. The semantics of OLP-FOL is a generalisation of the well-founded semantics. An OLP-FOL theory T = (T_d, T_c), divides the set of predicate symbols in two disjoint subsets: the defined predicates, which occur in the head of a clause of T_d, and the open predicates, which occur at the most in the body of the clauses of T_d.
In previous work, the composition of two OLP-FOL theories, with non-intersecting sets of defined predicate symbols, was studied. It was argued that their composition is given by the set of common models. Here, we investigate the possibility of composing two OLP-FOL theories, which define the same predicate. Therefore, we introduce two operators on theories: the p-opening operator, which opens the definition of the predicate p in a theory completely, and the conditional p-opening operator, which maintains the definition of p in a theory if a certain condition holds, and opens p in the other cases. We show that we can compose two theories, which both have an open definition for the same predicate, or which both have a conditional open definition for the same predicate, with non-overlapping conditions.
A functional-logic alternative to monads
Rafael Caballero-Roldan, Francisco J. Lopez-Fraguas
(Universidad Computense de Madrid, Spain)

Monads are a technique widely used in functional programming languages to address many different problems. This paper presents extensions, a functional-logic programming technique that constitutes an alternative to monads in several situations. Extensions permit the definition of easily reusable functions in the same way as monads, but are based on simpler concepts taken from logic programming, and hence they lead to more appealing and natural definitions of types and functions. Moreover, extensions are compatible with interesting features typical of logic programming, like multiple modes of use, while monads are not. This property further contributes to the employment of the same code to solve different problems.
Towards a Game-based Architecture for developing Complex Interactive Components in Computational Logic
K. Stathis
(Imperial College, London UK)
We present a game-based architecture for developing complex interactive components in computational logic. This supports the component oriented development of interactive systems, in such a way, that interactive components can be represented either as players that make moves according to the rules of a game or as umpires that enforce the rules on players, and thereby controlling their interactions. The architecture has already been successfully applied to develop knowledge-based front-ends to existing software components, including multi-lingual support for such applications. The potential of the architecture is that centralized organizations controlled by umpires or decentralized organizations in terms of autonomous players can be combined to form interactive systems of a very complex nature.
Distribution in a Demand Driven Style
Ian Holyer, Neil Davies and Eleni Spiliopoulou
(University of Bristol, UK)
Many application programs are now distributed. Despite the number of distribution models, they still suffer from great complexity and therefore lack resilience and trustability, mainly due to the non-deterministic underlying communication primitives. In this paper we present a demand-driven computational model which allows for arbitrary distribution of computation. This is based on deterministic concurrency, a purely declarative form of concurrency which extends the demand driven model of execution of functional languages into one with multiple independent demands and does not allow non-determinism to appear at the user level. The deterministic concurrency approach to distribution leads to a model where the issue of distribution is orthogonal to the issue of functionality. Its main characteristic is the high level of trustability in the presence of distribution which derives from the preservation of referential transparency. The declarative semantics of a program are not affected, no matter how the program is distributed. Furthermore, within this framework, all entities are mobile. Both code and data can be moved dynamically, on demand.
Using a Modular Distributed Temporal Logic for In-the-large Object Specification
J. Kuester Filipe
(Technical University Braunschweig, Germany)
Our general goal is to provide a semantic foundation for in-the-large specification of distributed information systems. We use TROLL, an object-oriented formal language, for system specification. Our claim is that objects are not enough as a modularisation unit when it comes to deal with very large systems. An intermediary concept between the system and the objects is needed for allowing reusability of specifications and provide a clearer system structure. Enriching TROLL with a module concept enforces us to develop new theoretical constructs ensuring an appropriate underpinning of the language. A modular distributed temporal logic MDTL has been developed which describes the dynamic aspects of modular systems. The main features of the logic are its own modularity, the ability to express inter-module (a)synchronous communication and intra-module concurrency. It also seems promising to support module refinement. MDTL is based on n-agent logics and interpreted over labelled prime event structures. In this paper we present the logic and semantics underlying the modular object-oriented specification language by means of a small toy example.
Composing Reusable Synthesis Methods through Viewpoints
Jutta Eusterbrock
(University of Aberdeen, Scotland, UK)
To make sense of research on constructive knowledge-based me\-thods for synthesis tasks, it has to be shown how to reuse and adopt these results in realistic settings. Cooperative reuse-oriented heterogenous systems have been proposed as a solution for achieving high-quality user-friendly systems within changing and distributed environments. This causes the question, how heterogenous knowledge resources, application systems and various kinds of synthesis knowledge, whose informal specifications don`t match with structural requirements for logic-based constructive reasoning, could be accessed and used by knowledge-based applications such that logical properties are being preserved. {\em Viewpoints} are intended to provide intermediate user-de\-finable independent components between reasoning components and the domain resources of interest. The starting point is given by a logical theory, which provides a generic description of a class of problems by specifying generic actions, goals and relationships. The definitions serve as initial elements of a task-specific library of synthesis me\-thods which are derived as meta-level theorems. A viewpoint is a commitment to represent domain specific knowledge fragments and problem-solving knowledge in terms of the vocabulary of an associated generic theory, such that correctness axioms are preserved. As a major point, it is proposed to share abstract structure through graphs rather than to use translations for knowledge integration. It will be claimed that the viewpoint concept is an adequate means for the representation of multiple perspectives on distinct semi-structured domain-specific knowledge resources, making them reusable for various synthesis applications and synthesis me\-thods reusable in various settings.
On Specification and Correctness of OOD Frameworks in Computational Logic
Kung-Kiu Lau (University of Manchester, UK)
Mario Ornaghi (Universita' degli studi di Milano, Italy)
In current component-based software development (CBD), it is widely recognised that the distribution of tasks between objects and the contracts between them are key to effective design. In composing designs from reusable parts, increasingly the parts are Object-oriented Design (OOD) frameworks, namely descriptions of the interactive relationships between objects which participate in the interactions. Designs are then built by composing these frameworks, and any object in the final design will play (various) roles from several frameworks. In this paper, we discuss our preliminary efforts to define a formal semantics in computational logic for the specification and correctness of OOD frameworks, and briefly illustrate it with frameworks in the CBD methodology Catalysis. The novelty of our approach is a priori correctness for OOD frameworks (and components in general) and their composition, in contrast to current development methods which are mainly in the style of posit-and-prove, whereby proof of correctness is done by a posteriori verification. For component-based software development, we argue that a priori correctness is a better approach than a posteriori correctness.
Back to COCL home page