Delta-Oriented Product Prioritization for Similarity-Based Product-Line Testing

Abstract

Testing every product of a software product line (SPL) is often not feasible due to the exponential number of products in the number of features. Thus, the order in which products are tested matters, because it can increase the early rate of fault detection. Several approaches have been proposed to prioritize products based on configuration similarity. However, current approaches are oblivious to solution-space differences among products, because they consider only problem-space information. With delta modeling, we incorporate solution-space information in product prioritization to improve the effectiveness of SPL testing. Deltas capture the differences between products facilitating the reasoning about product similarity. As a result, we select the most dissimilar product to the previously tested ones, in terms of deltas, to be tested next. We evaluate the effectiveness of our approach using an SPL from the automotive domain showing an improvement in the effectiveness of SPL testing.