Search

Aquest treball de final de grau té com a objectiu demostrar l'aplicació dels Models Conceptuals Deductius a través de la combinació de blockchain i bases de dades relacionals. La proposta dels MCD's va ser introduïda per Antoni Olivé l'any 1989, quan encara no exisitia la tecnologia blockchain. Aquesta tecnologia presenta un seguit de semblances amb la proposta d'Olivé, i en aquest projecte es demostra la possible implementació pràctica d'aquest concepte per a obtenir un sistema d'informació funcional. This final degree project has the purpose of demonstrating the application of the Deductive Conceptual Models through the combination of blockchain and relational databases. The Deductive Conceptual Models were introduced by Antoni Olivé on year 1989, before the blockchain technology was created. This very same technology presents multiple similarities with Olivé's proposal, so in this project we prove the possible practical implementation of this concept in order to accomplish a functional information system.

Dynamic integrity constraints (DIC) have been represented either as past-directed formulas or as future-directed formulas. This work, inscribed in the scope of the Deductive Conceptual Models of information systems, is an effort to prove the equivalence of these two kinds of formulas. Our work proposes a method that transforms any DIC specified through a future-directed formula into an equivalent past-directed formula. This method deals with the integrity constraints defined by a first-order logic language with explicit time that have the form of a denial. The method is based on the construction of time graphs to represent the temporal sequence of the facts presented in a formula. The first step consists in representing the initial formula defined in future tense by means of a time graph. Next a new time graph in a reverse temporal interpretation is constructed. Finally starting from this new graph the method generates a past-directed formula to represent the initial integrity constraint.

In this paper we present our approach for studying the temporal behaviour of a deductive conceptual model. It is based on the use of the semantic prototyping technique. We have implemented a tool able to simulate the execution of conceptual specifications and to give explanations about the state of the information base at any time, and about its temporal change and evolution. We believe that this tool can be useful in the important task of verifying the correctness of the conceptual specifications with respect to the user's requirements in a static as well as in a dynamic sense.

A new method for planning in a deductive conceptual model context is proposed. lt is based on viewing the information base modelled by a deductive conceptual model as a deductive database. This is possible provided that the database accomplishes sorne assumptions related to the time behaviour of predicates. Use of this plan generation method is useful for deductive conceptual models validation.

This paper describes the general framework and some details of the ODISSEA approach to conceptual modelling and design of information systems. We use a logic-based language for the specification of conceptual models and apply extensively logic-based techniques for the generation of a system design from a conceptual model. The main originality of our work is that our language is based on the deductive approach to conceptual modelling, instead of the usual operational approach. Deductive conceptual models (DCM) show a number of significant advantages, but are much more difficult to design and implement. We deal with the problem of generating a design for a given DCM. We present a new, formal method for generating transactions from a DCM. The method uses the SLDNF proof procedure and can be implemented in Prolog environments.

We propose an approach to validation of deductive conceptual models. This validation is performed through plan generation. The objective of plan generation is to determine sequences of events that lead to a certain goal state from a certain initial state. Our approach generates a plan in n stages obtaining at each of them the events of the plan that occur at each time point. At each stage, we use the intemal events model, obtained from the deductive conceptual model, which explicitly defines insertions and deletions of information induced by previous occurrence or absence of extemal events. An extension of SLDNF procedure can then be used to obtain sequences of externa! events that lead to the goal state. This work has been partially supported by PRONTIC program project TIC 680

In this paper we describe the ODISSEA language for the specification of information systems defined in the deductive approach. This language is being developed as part of the ODISSEA project, which aims at exploring the issues involved in the specification, design, construction, mantenance and evolution of what we call Deductive Information Systems. The sentences of the language are explained in detail with the aid of an example. We also give the complete BNF definition of the language. Finally, there is a brief explanation of the ODISSEA project and its execution and development environments.

We show that deductive conceptual models of information systems are a valid alternative to the usual operational models, and that they offer a number of advantatges. We describe the main features of DADES, as a particular example of deductive language. We also present the architecture of DADES/GP, an automatic prototype generator from models written in DADES.