A Hybrid Approach to Build Prioritized Pairwise Interaction Test Suites

Abstract

Traditional interaction testing aims to build test suites that cover all t-way interactions of inputs. But in many test scenarios, the entire test suites cannot be fully run due to the limited budget. Therefore it is necessary to take the importance of interactions into account and prioritize these tests of the test suite. In the paper, we use the hybrid approach to build prioritized pairwise interaction test suites (PITS). It adopts a one-test-at-a-time strategy to construct final test suites. But to generate a single test it firstly generates a candidate test and then applies a specific metaheuristic search strategy to enhance this test. Here we experiment four different metaheuristic search strategies. In the experiments, we compare our approach to weighted density algorithm (WDA). Meanwhile, we also analyze the effectiveness of four different search strategies and the effectiveness of the increasing iterations. Empirical results demonstrate the effectiveness of our proposed approach. Keywords-prioritized interaction test suites; greedy algorithm; metaheuristic search