Abstract. While recent data management technologies, e.g., object-oriented, address the problem of databases schema evolution, standard information systems currently in use raise challenging problems when evolution is concerned. This paper studies database evolution from the developer point of view. It shows how requirements changes are propagated to the database schemas, to the data and to the programs through a general strategy. This strategy requires the documentation of the database design. When absent, this documentation has to be rebuilt through reverse engineering techniques. The approach relies on a generic database model and on the transformational paradigm that states that database engineering processes can be modelled by schema transformations. Indeed, a transformation provides both structural and instance mappings that formally define how to modify database structures and contents. The paper then analyses the problem of program modification and describes a CASE tool that can assist developers in their task of system evolution. 1

Abstract. This paper describes an evolutionary approach to re-engineer a large medical imaging system using graph-transformations. The solution has been integrated into the ongoing development process within the organizational and cultural constraints of a productive industrial setting. We use graph transformations to model the original and the desired architecture, the mapping between architectural, implementation and design concepts, as well as related consistency rules. Violations are calculated every night and provided as continuous feedback to software architects and developers, so they can modify the system manually according to established procedures. With only a very limited global migration step and moderate changes and extensions to the existing procedures it was possible to improve the software architecture of the system, while new features still could be implemented and released in due time. Although this solution is dedicated to a concrete case, it is a starting point for a more generic approach. 1

Abstract. This paper presents a proposal that aims to cover some tasks required in the Database Reengineering process, mainly in the Conceptual Abstraction phase. The principal tasks are: the transformation of enriched logical schema to object-oriented schema, support for process iterations and support for data migration from the legacy database to the new database. The proposal consists of attaining the listed tasks by applying Formal Concept Analysis (FCA) techniques. It could be said that the advantage of using this proposal is that it is possible to build a tool, based on FCA, automatic, graphic and interactive. This tool will be able to support the process iterations, as it will be possible to mark the impact zones of later modifications in the logical schema and to verify how they affect the conceptual scheme obtained. This paper detailed presents the use of FCA in the Conceptual Abstraction phase. We obtain a class diagram that fits the object model defined for UML and which reflects the underlying model in the original logical schema. In the class diagram, we will have the significant classes of the problem, organized in the specialization/generalization hierarchy and related by means of associations (as a general framework for the rest of the UML relations, e.g. aggregations). The establishment of the associations can give rise to the appearance of new classes as association classes that characterize the defined relation.

Data migration, even though a practitioner’s topic, is addressed to good extent in academic literature. Nevertheless, there is a gap in literature regarding data migration under the influence of the specific circumstances of software product companies. The majority of publications address data migration from a perspective of a service provider that builds a onetime solution for a data owner. When software product companies re-engineer a data intensive product, a data migration solution is needed. However, the situation is different than that of classical data migration because the company has in house knowledge of the source, is itself building the target and must provide replicable deployment to all customers. Thus, software product companies are left to tailor data migration to a software product situation without any guidance from academic literature. To help software product companies manage data migration, this work proposes a software product data migration method. The method is built using assembly based situational method engineering. In order to use this technique existing literature is studied to identify methods, phases and activities of data migration. Situational method engineering is used to analyze and compare the existing methods in order to assemble a data migration method that includes a software product company perspective. The method is then evaluated by 11 data migration experts. The evaluation yielded positive results in terms of validity measures and allowed for the method to be refined based on comments and feedback from the experts. Finally, a case study on implementing the method at AFAS Software B.V., a Dutch software product company, is performed to attest the applicability of the method and bring to light insights regarding software product data migration.