Téléchargez les slides d’Isabelle Boydens (1 MB) et d’Anthony Cleve (30 MB).
La prochaine réunion du groupe de contact FNRS « Analyse critique et amélioration de la qualité de l’information numérique » se tiendra le mercredi 11 mai 2016 à 14h00 à l’Université Libre de Bruxelles (auditoire AY2.108, bâtiment A, campus du Solbosch).
Pluridisciplinaire, le groupe se situe au confluent des sciences appliquées et des sciences humaines et politiques.
Le groupe, dont nous avons fêté les vingt ans lors de la journée d’étude de janvier 2014, et qui s’est réuni ensuite en 2015 abordera cette année le thème suivant : « Analyzing the evolution history of data-intensive systems in support to software maintenance ».
Cette thématique nous conduira dans les méandres complexes de l’archéologie des données, domaine stratégique pour la qualité de l’information. En effet, les coûts de maintenance et de « reverse engineering » des bases de données en relation avec le code associé vont croissant dans la plupart des services informatiques de par le monde. L’exposé sera illustré de nombreux exemples, dont une vaste étude de cas dans le domaine des soins de santé au Canada.
La réunion se terminera par une table ronde au cours de laquelle tous les participants qui le souhaitent seront invités à intervenir et sera suivie d’une réception.
L’accès à la rencontre, qui est financée par le Fonds National de la Recherche Scientifique, est gratuit ; il est toutefois indispensable de s’inscrire avant le 8 mai 2016 au plus tard à l’événement, le nombre de places disponibles étant limité. Accédez au formulaire d’inscription.
Programme
13h30 Accueil des participants
14h00 Mot d’accueil par Françoise D’Hautcourt, Docteur en Sciences Appliquées, Professeur de l’ULB et Professeur honoraire, Membre fondatrice de la filière STIC (ex Infodoc), Membre fondatrice et première Présidente de ce groupe de contact FNRS
14h10 Introduction, par Isabelle Boydens, Présidente du groupe de contact FNRS « Analyse critique et amélioration de la qualité de l’information numérique », Professeur ordinaire à l’ULB et responsable du « Data Quality Competence Center » au sein du département Recherche de Smals
14h20 Analyzing the evolution history of data-intensive systems in support to software maintenance, par Anthony Cleve, Secrétaire du groupe de contact FNRS « Analyse critique et amélioration de la qualité de l’information numérique », Chargé de cours à la Faculté d’Informatique de l’Université de Namur et Maître de conférences au sein du MaSTIC de l’ULB
15h30 Débat et table ronde. Modérateurs : Joël Goossens, Chargé de cours au département d’informatique, service qualité et sécurité des systèmes informatiques, Faculté des Sciences, ainsi qu’au sein de la filière Mastic, ULB et Isabelle Boydens
16h00 Réception
Résumé
Analyzing the evolution history of data-intensive systems in support to software maintenance
Nowadays, software systems are ubiquitous, and a large part of our modern life deeply depends on them. Yet they are considered as one of the most complex artefacts ever built by human beings. In particular, maintaining such systems has long been recognized as a crucial, complex and costly task, which is especially due to the lack of sufficient documentation. In this context, understanding the software system constitutes a prerequisite of the maintenance and evolution processes. This understanding step, also called reverse engineering, may represent up to 50% of the total maintenance effort.
It has also been shown that understanding the evolution history of a complex software system can significantly aid the reverse engineering step and inform future system maintenance and evolution initiatives. Software repositories such as version management systems provide excellent opportunities for historical analyses. Most research work in this area has been concentrated on program code, design and architecture, but little attention has been devoted to the database of the system. This is an unfortunate gap as many software systems are data-intensive, i.e., their central artefact is a database.
Understanding the database schema which captures domain-specific concepts, data structures and integrity constraints may constitute a prerequisite to understanding the source code of the system as well as its evolution. Similarly, determining which fragments of the source code access certain parts of the database may be of crucial importance in the context of various software evolution scenarios such as migrating the application towards a new database platform, evolving the data structures to meet changing requirements, or assessing the quality of the overall system.
The goal of our current research effort is to reduce this gap by considering the evolution history of the database and its manipulation by the source code as an additional information source to inform software evolution tasks. In this talk, we will summarize our recent achievements towards this goal. In particular, we will present DAHLIA (Database ScHema EvoLutIon Analysis), a visual analyzer of database schema evolution and manipulation. This tool mines the database schema evolution history from the software repository and allows its interactive, visual analysis.
Through a sophisticated static analysis approach, DAHLIA also allows developers to identify and analyse database accesses in highly dynamic systems and to formulate recommendations about the impact of database schema changes on the programs. We will then report on the use of DAHLIA to analyse large-scale data-intensive systems including OSCAR, a large open-source information system that is widely used in the healthcare industry in Canada.
Credits:
This talk will be about joint work with the following fantastic researchers:
- – Loup Meurice, Maxime Gobert and Jérôme Maes (University of Namur, Belgium)
- – Csaba Nagy (University of Szeged, Hungary)
- – Jens Weber (University of Victoria, Canada)
Intervenant
Anthony Cleve is a professor in information system evolution at the University of Namur, where he is the director of the PReCISE research center, bringing together over 50 full-time researchers, focused on the engineering and management of advanced information systems. He is also a visiting Lecturer in database engineering at the Université libre de Bruxelles. His research interests include data-intensive system maintenance and evolution, software and data reengineering & migration, data modeling, program analysis & transformation, and self-adaptive and context-aware systems.He has co-authored over 50 publications related to software and systems engineering, published in international conferences and journals.
Anthony’s PhD thesis, entitled « Program Analysis and Transformation for Data-Intensive System Evolution », earned him the IBM Belgium 2010 Award for the best PhD thesis in Computer Science and applications. He was also nominated for the Cor Baayen Award 2012, offered to a promising young researcher in computer science and applied mathematics. Anthony regularly serves in the organizing and/or program committees of many international scientific events. He served as co-chair of the ERCIM Working Group on Software Evolution, Steering Committee member of the International Workshop on Principles of Software Evolution, and program co-chair for the 2012 and 2013 editions of the European Conference on Software Maintenance and Evolution.