LASIGE’s PhD student Allan da S. Espindola and integrated members António Casimiro and Pedro M. Ferreira in collaboration with Altair O. Santin and Eduardo K. Viegas from the Pontifical Catholic University of Paraná (PUCPR), Brazil, was published in the top 10% journal *Future Generation Computer Systems*. The paper entitled “Enhancing Intrusion Detection Generalization via Diversity-Driven Multi-View Ensemble Learning in Industrial Systems” addresses the challenge of detecting previously unseen attacks in Supervisory Control and Data Acquisition (SCADA) systems, which are used to monitor and control critical industrial infrastructures.

The study introduces DIME-IDS, a diversity-driven multi-view ensemble that combines four complementary views: network, host, user activity, and system activity. Bringing together evidence from these views gives the intrusion detection system broader visibility. In the experiments, DIME-IDS detected unseen attack behaviors more reliably and produced fewer false negatives than the single-view and concatenated approaches evaluated by the authors.

The paper is available here.