Leveraging relational concept analysis for automated feature location in software product lines

Abstract

Formal Concept Analysis (FCA) has been introduced for almost a decade as a suitable method for Feature Location (FL) on a collection of product variants. Even though FCA-based FL techniques allow to locate the core of a feature implementation, they do not propose a solution to trace feature interactions to their implementation. Thus, the extracted traceability links (traces) are too inaccurate, and, in the context of SPL extraction, cannot be used to generate complete products. In this paper, we propose to complement the FCA-based FL techniques by leveraging the power of Relational Concept Analysis, an extension of FCA to multi-relational data. From two given formal contexts, one for the product's artefact and one for their features, our technique computes the traces that link the features and the feature interactions to their corresponding artefacts. Additionally, we introduce a stage that removes unnecessary features from the extracted traces, to make them easier to understand by an expert. Our FL technique can be applied at any artefact granularity (from files to statements) and independently from software languages. The results show that our technique produces valid traces, from which we were able to completely rebuild the set of artefacts for each initial product. Moreover, they show that our trace reduction removes, on average, between 31% and 85% of unnecessary features from the traces. Keywords feature location • software product variants • software product lines • formal concept analysis • relational concept analysis * Citation: Hlad et al. Leveraging Relational Concept Analysis for Automated Feature Location in Software Product Lines.

Relational Concept Analysis in Practice: Capitalizing on Data Modeling Using Design Patterns

“…Finally, RCA was applied in Software Engineering in order to solve different problems: normalize (by factorization) a UML class model [19,16] or a UML use case model [10]; for refactoring [26]; learn model transformation patterns [9]; build directories of Web services or components [5]; structure [7] or analyze [2,18] the variability in software product lines.…”

“…helping identifying which ones are equivalent on all option categories, how they differ from each other, and how the options of the different categories interact. Moreover, an application has been developed by [18] using this DP to assist Feature location (FL) in Software Product Line Engineering.…”

“…To improve the readability of the relational attributes, which is a drawback of RCA, the simplification proposed by [33,34] can be applied: a concept identifier in a relational attribute can be replaced by the intent of the concept at its creation time (possibly recursively). Other authors proposed to deliver the extracted knowledge, not in terms of concepts and relational attributes, but as sentences [29] or as logical formulas in source code macros [18].…”

Detecting and Visualizing Implementation Feature Interactions in Extracted Core Assets of Software Product Line

“…Otherwise, the performance of the generated products will be degraded, and unexpected behavior may occur. This is known as unwanted feature interactions [13], and it occurs when multiple feature implementations are combined in a new product, and their behaviors are unexpected even if the implementation of each individual feature is working correctly and independent in their domain. This kind of feature interaction at the source code level is known in the literature as structural interaction or implementation dependency [19] [17].…”

“…The implementations of extracted features from product variants are overlapped in some classes or methods as features interact in software [19] [13] [10]. These overlapped artifacts (shared artefacts) do not represent the core implementations of features but they are added to allow two or more features to work as a combination in their hosted software variants.…”