The Generation of Optimized Test Data: Preliminary Analysis of a Systematic Mapping Study

Abstract

Test data generation for algorithm testing is one of the widely studied topics in software testing researches. Although there are various well-known approaches based on optimization methods, which aim to generate test data such as Simulated Annealing, Ant Colony and Genetic Algorithms, there is no systematic study, which classifies these approaches according to special requirements such as comparing the novelty of the proposed approach with the well-known methods and the types of the benefits, if any. The objective of this paper is to provide an information to assist and guide researchers on this research area by supporting further research efforts. In order to close the gap in the existing literature and address this need, we have already started a systematic mapping study. In this paper, we present the preliminary analysis of this ongoing study with a subset of overall pool, which includes 2635 papers. More specifically, in this paper, we examined ~10% of our final pool of papers (i.e., 260 papers) and found 42 relevant studies, which addresses our research questions. The preliminary analysis, showed that the studies, which focus on generating optimized test data, either create new approaches or improve the existing approaches by adding new features. Moreover, our preliminary results also showed that there is an open research area in this topic.