Résumé :

As the volume of data continues to grow, the complexity of database systems has surged, giving rise to NoSQL systems. This forces Information Systems (IS) architects to constantly adapt their data models (i.e., the structure of stored information) and meticulously select the optimal option(s) for data storage and management. In response to this challenge, we propose an automated global approach for guiding the transformation of data models. Our approach initiates by transforming the conceptual model into a logical data model, then applies refinement rules recursively to generate all possible data models. To reduce the search space, our approach employs a heuristic that avoids redundancies and considers the use case. Additionally, we propose a cost model, enabling the comparison of generated data models at a logical level. This cost model integrates both data model and query costs, along with considerations for environmental impact, financial, and time costs. For the first time, our approach introduces a multidimensional cost model that evaluates time, environmental, and financial constraints. This cost model eases the comparison of data models, helping with the selection of the most suitable one for a given use case and context by optimizing either the environmental impact or the stability of the data model.