Exploring Functionality and Efficiency of Feature Model Product Configuration Solutions

• Published in: IEEE access Vol. 10; p. 1
• Main Authors: Vidal-Silva, Cristian, Cardenas-Cobo, Jesennia, Sanchez, Aurora, Duarte, Vannessa, Tupac-Yupanqui, Miguel
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.01.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: AAFM; AAFM solutions; Automated Analysis of Feature Model; Automation; Configurations; conflict detection; Content management software; Content management systems; Diagnosis; Feature model; Linux; product completion; product configuration; Python; speculative programming
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3231449

Variability-intensive systems are software systems in which variability management is a core activity. Examples of variability-intensive systems are the web content management system Drupal, the Linux kernel, and the Linux Debian distributions. Feature models have been considered useful tools for modeling variability-intensive systems for more than 30 years, and their automated analysis is a thriving, motivating, and active research area. In 2010, Benavides et al. published the survey results of the first 20 years of Automated Analysis of Feature Model solutions. At that time, mainly sequential computing solutions exist. The product configuration of feature models represents a relevant operation demanding efficient automated solutions, which are now possible for assisting the feature model product configuration, such as minimal conflict detection, diagnosis, and product completion. This article reviews the fundaments of product configuration of feature models and current solutions for the Automated Analysis of Feature Model. It assesses the functionality and computing performance of commonly used Automated Analysis of Feature Model solutions for minimal conflict detection, minimal diagnosis, and the minimal completion of partial product configuration and the approaches. This article highlights research opportunities for developing new and more efficient solutions for conflict detection, diagnosis, and product completion of large-scale configurations. We describe a promising computing approach for that purpose.