Sequence diagram to colored Petri nets transformation testing

Abstract

Model transformation is an important concern in software engineering today and rigorous techniques are needed to check transformation quality. In this paper, we propose a novel oracle function for model transformation testing that is based on the notion that the more a transformation deviates from well-known transformation examples, the more likely it is in error. Our approach takes its inspiration from the biological metaphor of immune systems, where pathogens are identified by their difference with normal body cells. We also propose a visualization module to explore transformation fragments according to their risk. We evaluate our approach with a complex transformation mechanism: Sequence Diagram to Colored Petri Nets. The validation results confirm the effectiveness and efficiency of our approach.

Model Transformation Testing and Debugging: A Survey

“…We distinguish the following categories. MC-E M2M Any E-M Wieber et al [131] MC-E M2M Any E-A Lin et al [76] MC-E M2M ECL E-M Tiso et al [117] MC-E M2T Acceleo E-M Stürmer et al [113] MC-S M2T TargetLink E-A Finot et al [38,81] MC-S M2M Any E-M Kolovos [71] MC-S M2M ECL E-M Kessentini et al [68,69] BE/TB M2M Kermeta E-A Matragkas et al [77] BE/TB M2M ETL Tr-A Jörges and Stefen [66] BE/TB M2T CG Tr-A Metamorphic testing. This is a technique to alleviate the oracle problem [184,242,243].…”

“…We identify three types of approaches in this category. First, Kessentini et al [68,69] present an oracle function based on the notion that the more an MT deviates from well-known MT examples, the more likely it is to be faulty. They compare the output model not with a corresponding expected output for the given input model, but with an already available set of examples which contain good quality MT traces from past MTs.…”

Utilizing Automatic Query Reformulations as Genetic Operations to Improve Feature Location in Software Models

“…Outside the SBMDE community, there are many crossover and mutation operations proposed [16], [17], [18]. These surveys identify more than 50 crossover operations [20] -Random Kessentini et al [21] -Random Faunes et al [22], [23] N point Random Baki et al [24] Single-point Random Saada et al [5] Single-point Random Kessentini et al [25] -Random Mkaouer et al [26] Single-point Random Gyapay et al [27] --Denil et al [3] -Random Abdeen et al [28] Single-point Random Fleck et al [29] Single-point Random Nisbet et al [30] --Cooper et al [31] Single-point Random Fatiregun et al [32] --Fatiregun et al [33] Single-point Random Kulkarni et al [34] Single-point Random Fatiregun et al [35] Single-point Random Fleurey et al [4] -Random Cadavid et al [36] -Random Wang et al [37] -Random Shelburg et al [38] -Random Jilani et al [39] -Random Sahin et al [40] Single-point Random Rose and Poulding [41] -Random Batot [42] Single-point Random Kessentini et al [43], [44] Single-point Random Faunes et al [45] Single-point Random Kessentini et al [46] Single-point Random Mansoor et al [47] Single-point Random Debreceni et al [48] --Ghannem et al [49] Single-point Random Ghannem et al [50] Single-point Random Amoui et al [51] Single-point Random Jensen and Cheng …”