Yamine Aït Ameur

An a Priori Approach for Automatic Integration of Heterogeneous and Autonomous Databases

Data integration is the process that gives users access to multiple data sources through queries against a global schema. The semantic heterogeneity has been identified as the most important and toughest problem when integrating various sources. The mapping between the global schema and local schemas was done manually in the first generation of integrated systems, when ontologies are not used to make explicit data meaning. It is semi automatic when ontologies and ontology mappings are defined at integration level. In this paper, we propose a fully automatic integration approach based on ontologies. It supposes that each data source contains a formal ontology that references a shared ontology. The relationships between each local ontology and the shared ontology are defined at the database design time and also embedded in each source. We assume that a domain ontology exists, but each source may extend it by adding new concepts and properties. Our approach is currently prototyped in various environments: OODB, ORDB, and RDB.

Using the B Formal Approach for Incremental Specification Design of Interactiv Systems

This paper introduces a new technique for the verification of interactive systems. It first presents the use of a model oriented formal method for specifying interactive systems, i.e. the B method. Then, it suggests formally based solutions which allow solving difficulties that are inherent to interactive systems specification, like reachability, observability or reliability. Our claim is that this model-oriented technique that uses proof obligations can be used together with model checking techniques, where automatic proofs of properties can be performed.

Bases de données à base ontologique. Principe et mise en oeuvre

In this paper, we propose a new database model called Ontology-Based Database (OBDB) that has two main characteristics. First, it allows to manage both ontologies and data. Second, it permits to associate each data with the ontological concept which defines its meaning. Our approach addresses ontologies expressed as models, i.e. sets of objects that may be accessed from a root object, and that consists of classes and properties. Target data are those data that are instances of ontology classes and that fulfil a strong typing assumption: each instance shall belong to exactly one base class, defined as the lower bound (for the subsumption relationship) of the classes the instance belongs to, and it shall be described only by properties that are applicable to its base class. We first propose an OBDB architecture model, called OntoDB. Then we outline its implementation using the PostgreSQL DBMS, the PLIB ontology model and the EXPRESS language.

PLIB Ontology: A Mature Solution for Products Characterization in B2B Electronic Commerce

This paper deals with the integration of PLIB ontologies in electronic commerce architectures, especially in B2B electronic commerce. Exchange of data, information, and knowledge is the key issue. To ensure reliability and feasibility of such exchange, two kinds of ontologies prove useful. One addresses products and services characterization, whereas the other addresses business processes. This paper aims to demonstrate the capability of integration of these two kinds of ontologies in a unique and global B2B-layered architecture. The orthogonality principle is applied. It allows the use of any best existing standard in both business layer and contents layer. In this paper, we use RosettaNet PIPs and ebXML in the business process layer, and we claim that PLIB ontologies are the best-in-class solution for the contents layer. These standards are used conjointly to promote interoperability and automation between partners involved in exchange. The originality of our approach is the possibility offered to the partners to use any defined business process with any defined content. A case study showing an application of such ontologies using RosettaNet PIP2A9 for business process and PLIB-based ontology for product characterization is presented. Implementation issues are also discussed.

Preface: Special issue on contribution of ontologies in designing advanced information systems

This special issue on contribution of ontologies in advanced information systems was initially discussed and plannedwith Prof. Reind van de Riet. Note that it is unusual to organize a special issue for Data & Knowledge Engineering Journal without a supporting conference or workshop. We got this opportunity after a long discussion and exchanges with Prof. Reind van de Riet. After he passed away, Prof Peter Chen contributed to the finalization of this issue. The subject of our special issue is the ontologies and their contribution in advanced information systems. Information Systems are repositories where sensitive and crucial data are recorded to support day-to-day company applications and decision making processes. Therefore, these systems often contain most of the company product and process knowledge. Unfortunately, this knowledge is implicitly encoded using company specific models. Thus, developing new user interfaces or reconciling data with external ones often require some kind of reverse engineering processes for making data semantic explicit. Nowadays, ontologies are used for making explicit the meaning of information in several research and application domains. Initiated in Artificial Intelligence, ontologies are now used in a large spectrum of fields such as: Semantic Web, information integration, database design, e-Business, data warehousing, and data mining. They are also used to provide information system user knowledge-level interfaces. Over the last 5 years, a number of interactions between ontologies and information systems have emerged. Newmethods have been proposed to embed within database of both ontologies and data, defining new ontology-based database. New languages were developed in order to facilitate exchange of both ontology and data. Other languages dedicated for querying data at the ontological level were proposed (e.g., RQL, Sparql, OntoQL). Various approaches have been designed to deal with semantic integration of heterogeneous information sources using ontologies either in data sources or in mediators. In some domains, like product modeling, ontologies were published as standards. These ontologies are actually used to define world wide exchange consortiums for sharing information in various application domains. Due to these recent developments, most commercial database systems offer solution to both manage ontologies and data (e.g., Oracle, IBM Sor). The aim of this special issue of the Data & Knowledge Engineering Journal is twofold. The first one is to present new and challenging issues in the contribution of ontologies for designing high quality information systems. The second one is to present new research and technological developments that use ontologies all over the life cycle of information systems. There was a great response to the call for papers; we received 128 abstracts, and finally received 65 papers from 22 countries. Due to this success and the quality of the received papers, Prof. Chen gave us the opportunity to accept 8 papers for this special issue which makes an accepting rate of 12.30%. These papers are authored by an outstanding roster of experts in their respective fields, and tackle various issues from different angles, requirements and interests. Their topics include: ontology-based system integration, ontology-driven information systems, ontology-enhanced information systems, reasoning, semantic data warehousing, case tools, semantic conflict detection, ontologies and data mining. The eight selected papers are summarized as follows:

Keynote 1: Extending Persistent Meta-Modeling Systems to Handle Behavioral Semantics

Summary form only given. The increasing number of information systems modelling formalisms raises several issues related to data and models integration or exchange. To address these problems, several meta-modelling systems have been proposed. However, few of them use a database as a back-end repository in order to offer a persistent solution for addressing over sized models. The aim of this talk is to show how the benefits of classical modelling systems, typically their capability to express behavioural semantics of models elements combined with the advantages of databases, i.e. their scalability and querying capabilities make it possible to support behavioural descriptions within PMMS. The proposed approach focuses on the capability to dynamically introduce new operators that could be implemented with an external program stored outside the database, or with a web service, etc. As a consequence of this extension, model transformations in model-based repositories, computation of derived concepts, web services triggering from relational databases, information exchange and data integration become activities supported in a persistent context.

Formal program developments directed by operational properties evaluation

Transformation systems allow to support developments leading from an initial specification to a final program in a totally formal way. Transformations use valid properties of input objects to produce new equivalent ones. Most transformations use functional properties to increase the efficiency of programs. Doing so, they affect non nonfunctional properties which, more often, are not formally expressed. The problem the author addresses is to recognize situations in which transformations can be achieved on the basis of the evaluation of a defined nonfunctional property and his aim is to relate program transformations to nonfunctional properties evaluations. Indeed, a particular transformation tactic can be applied when a given property does not hold. The DEVA language has been used as a support of experiments in the development of programs.<<ETX>>