Automated generation of computationally hard feature models using evolutionary algorithms
作者：
Highlights：
• A novel evolutionary algorithm for the generation of Feature Models (FMs) is presented, named ETHOM.
• ETHOM is used to generate hard FMs that show the performance of FM analysis tools in pessimistic cases.
• ETHOM contributes to assess the performance of analysis tools, revealing bugs and driving their optimization.
• ETHOM generates FMs extremely harder to analyse than random models of identical or even larger size.
摘要
•A novel evolutionary algorithm for the generation of Feature Models (FMs) is presented, named ETHOM.•ETHOM is used to generate hard FMs that show the performance of FM analysis tools in pessimistic cases.•ETHOM contributes to assess the performance of analysis tools, revealing bugs and driving their optimization.•ETHOM generates FMs extremely harder to analyse than random models of identical or even larger size.
论文关键词：Search-based testing,Software product lines,Evolutionary algorithms,Feature models,Performance testing,Automated analysis
论文评审过程：Available online 27 December 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.12.028
