Ángel Jesús Varela-Vaca

Feature models (FMs) play a crucial role in software product lines (SPLs) by representing variability and enabling the generation of diverse product configurations. However, the vast number of possible configurations often makes it challenging to identify the most suitable variant, especially when multiple criteria must be considered. Multi-criteria decision-making (MCDM) methods, such as analytic hierarchy process (AHP), technique for order of preference by similarity to ideal solution (TOPSIS), and VIseKriterijumska Optimizacija I Kompromisno Resenje (“multicriteria optimization and compromise solution”) (VIKOR), are effective for ranking configurations based on user-defined preferences. However, the application of disparate MCDM techniques to the same feature model with identical criteria can yield conflicting rankings, thereby complicating the decision-making process. To address this issue, we propose a novel framework that systematically integrates multiple MCDM methods to prioritise product configurations and provides informed decision support to reconcile ranking discrepancies. The framework automates the prioritisation process and offers a structured approach to explain differences between rankings, enhancing transparency and user confidence in the final selection. The framework’s effectiveness has been validated through real-world case studies, demonstrating its ability to streamline configuration prioritisation and support consistent, preference-driven decision-making in complex SPL environments.

The proliferation of connected smart homes presents a conundrum to society - security versus convenience. As organisations invest heavily in developing products and services to automate domestic tasks, consumers are increasingly being lured towards unsecured devices that often offer enticing and immediate benefits. These devices range from security cameras to climate control systems, intelligent lighting, and more. The evolution of artificial intelligence offers a potential answer to identifying cyberattacks and vulnerabilities in smart home environments. This study intends to establish the existing architectures, frameworks, techniques, data sources and environments to which the current state of the art has contributed. To that extent, a systematic literature review (SLR) process has been carried out across four journal databases - SCOPUS, ACM, IEEE Xplore and Science Direct. 390 papers were identified in the selection stage, 131 progressed to the extraction stage, where 90 were finally accepted. The findings show that significant efforts have been made to test the detection accuracy of various artificial intelligence techniques using standard open datasets of known attacks, or simple datasets from prototype testbeds. Attempts were also made to identify zero-day vulnerabilities through anomaly detection techniques, although the breadth and depth of the datasets employed are limited. The summary of this review reveals the gaps across datasets, techniques, and missing external data context from sources, such as the National Vulnerability Database (NVD), amongst others. These present promising opportunities for research in the field.