Selection hyper-heuristic for software testing

• Grant number: 18/08372-8
• Support Opportunities: Scholarships abroad - Research
• Start date: January 02, 2019
• End date: January 01, 2020
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Valdivino Alexandre de Santiago Júnior
• Grantee: Valdivino Alexandre de Santiago Júnior
• Host Investigator: Ender Ozcan
• Host Institution: Instituto Nacional de Pesquisas Espaciais (INPE). Ministério da Ciência, Tecnologia e Inovação (Brasil). São José dos Campos , SP, Brazil
• Institution abroad: University of Nottingham, University Park, England

Abstract

Search-Based Software Testing (SBST) has shown to be very promising for several activities of the testing process and its optimization as a whole. Recently, researchers have been addressing multi-objective problems and the use of hyper-heuristics. However, hyper-heuristics are still very incipient in the context of SBST where few works in the literature have been published with purposes such as Combinatorial Interaction Testing data generation, and integration and test order problem. Moreover, recent studies show that the most used selection hyper-heuristics within SBST, such as Choice Function, presented worse performance compared with other hyper-heuristics considered in different domains. This research project aims at investigating selection hyper-heuristics in the context of software testing. The idea is to create a new selection hyper-heuristic to choose Low-Level Heuristics (LLHs) based on approaches applied to domains such as wind farm layout optimization and scheduling, but which have never been used for software testing. Multi-Objective Evolutionary Algorithms will be considered as LLHs and the goal is to investigate classical selection methods (e.g. Fixed Sequence), and probabilistic models (e.g. Markov Chains). A rigorous evaluation, comparing the new methodology via selection hyper-heuristic with other approaches based on pure meta-heuristics, will be conducted with case studies, including space software systems, in a validation context still not investigated within testing based on hyper-heuristics. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DE SANTIAGO JUNIOR, VALDIVINO ALEXANDRE; OZCAN, ENDER; DE CARVALHO, VINICIUS RENAN. Hyper-Heuristics based on Reinforcement Learning, Balanced Heuristic Selection and Group Decision Acceptance. APPLIED SOFT COMPUTING, v. 97, p. 23-pg., 2020-12-01. (18/08372-8)

DE SANTIAGO JUNIOR, VALDIVINO ALEXANDRE; OZCAN, ENDER; DE CARVALHO, VINICIUS RENAN. Hyper-Heuristics based on Reinforcement Learning, Balanced Heuristic Selection and Group Decision Acceptance. APPLIED SOFT COMPUTING, v. 97, n. A, DEC 2020. (18/08372-8)

DE SANTIAGO JUNIOR, VALDIVINO ALEXANDRE; OZCAN, ENDER; BALERA, JULIANA MARINO. Many-objective test case generation for graphical user interface applications via search-based and model-based testing. EXPERT SYSTEMS WITH APPLICATIONS, v. 208, p. 21-pg., 2022-07-18. (18/08372-8)