André Miralles

Feature Model Composition Assisted by Formal Concept Analysis.

In the domain of software product lines, Feature Models (FM) play a central role in variability modeling, completed by configuration collections (from concrete software product lines), logical representations, constraint programming or conceptual structures, coming from the field of Formal Concept Analysis (FCA). The development of feature models may take several forms, including their synthesis from configuration collections or their design in several steps (by several teams or with different concerns). FM composition (merge) operators are part of that design activity as they assist their iterative building. In this paper, we describe an approach, based on two main merging semantics (intersection and union), which assists designers in merging several FMs. This approach benefits from the help of FCA to represent all the FMs with the same configuration set through a canonical form. We describe the implementation of our approach and present a set of concrete examples.

Application of a Model Transformation Paradigm in Agriculture: A Simple Environmental System Case Study

In this chapter, the authors use the methodology presented in Chapter 2 to develop a system that manages the spreading of organic waste on agricultural parcels. The proposed method uses a process of iterative and incremental development. Two complete iterations of the development process are presented starting from the analysis model and ending with the code produced by the case-tools SQL code generator. The first iteration deals with the description of territory objects and the second one deals with the business objects used in the context of the spreading of organic waste. As a result of transformations applied, models are enriched with new concepts and, therefore, are more complex. The growing complexity of the model may negatively affect an actors understanding, which may become an impediment by slowing down the analysis phase. The authors show how the software development process model, a modeling artifact associated with the continuous integration unified process method, avoids the apparent complexity of the model and improves productivity.

A New Methodology to Automate the Transformation of GIS Models in an Iterative Development Process

In the majority of research today in areas such as evaluation of flood risks, management of organic waste as it applies to plants, and mapping ecological conditions of rivers, scientific advances are often aimed toward the development of new software or the modification of existing software. One of the particulars for software developed for agricultural or environmental fields is that this software manages geographic information. The amount of geographic information has greatly increased over the past 20 years. Geographic Information Systems (GISs ) have been designed to store this information and use it to calculate indicators and to create maps to facilitate the presentation and the appropriation of the information. Often, the development of these GISs is a long and very hard process. Since the early 1970 s, in order to help project managers, software development processes have been designed and applied. These development processes have also been used for GIS developments. In this chapter, the authors present a new methodology to realize GIS more easily and more interactively. This methodology is based on model transformations, a concept introduced by the Object Management Group (OMG) in its approach called model driven architecture (MDA ). When software is developed, models are often used to improve the communication between users, stakeholders, and designers. The changes of a model can be seen as a process where each action (capture of user concepts, modification of concepts, removal of concepts, etc.) transforms the model. In the MDA approach, the OMG recommends automation of these actions using model transformations. The authors have developed a complete set of model transformations that enable one to ensure the evolution of a GIS model from the analysis phase to the implementation phase.

Sizing the Underlying Factorization Structure of a Class Model

The design of class models for information systems, databases or programming is a delicate process in which experts of the domain and designers have to identify and agree on the domain concepts. Formal Concept Analysis (FCA) has been proposed for supporting this collaborative work and fostering the emergence of higher level entities and the factorization of descriptions and behaviors. More recently, an extension of FCA, Relational Concept Analysis (RCA), has been designed to extend the scope of FCA to the emergence of higher level domain associations. FCA and RCA build a kind of normal form for models, in which the factorization is exhaustive, and the specialization order is adequate. The counterpart of these strong properties is a worst-case exponential theoretical complexity. In this paper, we study a practical application of RCA on several versions of a real class model in order to give precise figures about RCA and to detect which configurations are tractable.

Méthode de factorisation progressive pour accroître l’abstraction d’un modèle de classes

Nous nous interessons a la construction du modele de classes d’un systeme d’information environnemental dans un contexte multi-acteur. Nous etudions une amelioration progressive du modele par analyse des classes et des associations pour en faire emerger d’autres ayant un niveau d’abstraction plus eleve. La technique que nous utilisons est basee sur l’analyse formelle de concepts. Son application en une seule etape produisant de nouveaux concepts en nombre trop eleve pour l’analyse par les experts, nous proposons une construction par etapes. Cette solution est testee sur un modele de systeme d’information environnemental.

Assisting Configurations-Based Feature Model Composition

Feature Models (FMs) have been introduced in the domain of Software Product Lines (SPL) to model and represent product variability. They have become a de facto standard, based on a logical tree structure accompanied by textual cross-tree constraints. Other representations are: (product) configuration sets from concrete software product lines, logical representations, constraint programming, or conceptual structures, coming from the Formal Concept Analysis (FCA) framework. Modeling variability through FMs may consist in extracting them from configuration sets (namely, doing FM synthesis), or designing them in several steps potentially involving several teams with different concerns. FM composition is useful in this design activity as it may assist FM iterative building. In this paper, we describe an approach, based on a configuration set and focusing on two main composition semantics (union, intersection), to assist designers in FM composition. We also introduce an approximate intersection notion. FCA is used to represent, for a product family, all the FMs that have the same configuration set through a canonical form. The approach is able to take into account cross-tree constraints and FMs with different feature sets and tree structure, thus it lets the expert free of choosing a different ontological interpretation. We describe the implementation of our approach and we present a set of concrete examples.

Finding Semi-Automatically a Greatest Common Model Thanks to Formal Concept Analysis

Data integration and knowledge capitalization combine data and information coming from different data sources designed by different experts having different purposes. In this paper, we propose to assist the underlying model merging activity. For close models made by experts of various specialities on the same system, we partially automate the identification of a Greatest Common Model (GCM) which is composed of the common concepts (core-concepts) of the different models. Our methodology is based on Formal Concept Analysis which is a method of data analysis based on lattice theory. A decision tree allows to semi-automatically classify concepts from the concept lattices and assist the GCM extraction. We apply our approach on the EIS-Pesticide project, an environmental information system which aims at centralizing knowledge and information produced by different research teams.