An Ontological Rule-Based Approach for Analyzing Dead and False Optional Features in Feature Models

Luisa Rincón (1) , Gloria Lucia Giraldo (1) , Raúl Mazo (2) , Camille Salinesi (2)

1 Universidad Nacional de Colombia (COLOMBIA)
2 CRI - Centre de Recherche en Informatique de Paris 1

Product lines engineering uses Feature Models (FMs) as a notation to represent variability and commonality in families of products. One of the well-known issues of FMs is that they may have defects that can drastically diminish the benefits of the product line approach. Two of these defects are dead features and false optional features. Dead features are features absent from any valid product of the product line. False optional features are features declared as optional but actually required in all valid products. These two types of defects are undesirable in FMs because they give a wrong idea of domain that represents the FM. Several techniques documented in literature help to identify dead and false optional features. However, only few of them tackle the problem of identifying the causes of these defects. Besides, the explanations they provide are cumbersome and hard to understand by humans. In this paper, we propose an ontological rule-based approach to (i) identify dead and false optional features in FMs; (ii) identify certain causes of these defects; and (iii) explain these causes in natural language. Moreover, we propose a collection of rules that (i) formalize some cases that produce dead and false optional features; (ii) find the FM's elements that causes each defect; and (iii) explain why a feature is dead or false optional. This collection of rules helps modelers to correct the defects found in FMs and helps prevent the occurrence of new ones. We illustrate our approach in a reference model from literature. A preliminary empirical evaluation of our approach, using a benchmark composed of 31 FMs of sizes up to 150 features, shows that the proposal is effective, accurate and scalable.

Mots clés: Feature Models; Defects; Ontologies; Software Engineering

Domaines: Génie logiciel [cs.SE]; Intelligence artificielle [cs.AI]

Luisa Rincón, Gloria Lucia Giraldo, Raúl Mazo, Camille Salinesi. An Ontological Rule-Based Approach for Analyzing Dead and False Optional Features in Feature Models. XXXIX Latin American Computing Conference (CLEI), Oct 2013, Naiguatá, Venezuela. ⟨hal-00913944⟩