Adversarial attacks represent a critical issue that prevents the reliable integration of machine learning methods into cyber defense systems. Past work has shown that even proficient detectors are highly affected just by small perturbations to malicious samples, and that existing countermeasures are immature. We address this problem by presenting AppCon, an original approach to harden intrusion detectors against adversarial evasion attacks. Our proposal leverages the integration of ensemble learning to realistic network environments, by combining layers of detectors devoted to monitor the behavior of the applications employed by the organization. Our proposal is validated through extensive experiments performed in heterogeneous network settings simulating botnet detection scenarios, and consider detectors based on distinct machine- and deep-learning algorithms. The results demonstrate the effectiveness of AppCon in mitigating the dangerous threat of adversarial attacks in over 75% of the considered evasion attempts, while not being affected by the limitations of existing countermeasures, such as performance degradation in non-adversarial settings. For these reasons, our proposal represents a valuable contribution to the development of more secure cyber defense platforms.

This paper proposes a novel approach for the evaluation of the performance achieved by trainees involved in cyber security exercises implemented through modern cyber ranges. Our main contributions include: the definition of a distributed monitoring architecture for gathering relevant information about trainees activities; an algorithm for modeling the trainee activities using directed graphs; novel scoring algorithms, based on graph operations, that evaluate different aspects (speed, precision) of a trainee during an exercise. With respect to previous work, our proposal allows to measure exactly how fast a user is progressing towards an objective and where he does wrong. We highlight that this is currently not possible in the most popular cyber ranges.

Road traffic injuries are a critical public health challenge that requires valuable efforts for effective and sustainable prevention. Worldwide, an estimated 1.2 million people are killed in road crashes each year and as many as 50 million are injured. An analysis of data provided by authoritative sources can be a valuable source for understanding which are the most critical points on the road network. The aim of this paper is to discover data about road accidents in Italy and to provide useful visualization for improving road safety. Starting from the annual report of road accidents of the Automobile Club of Italy, we transform the original data into an RDF dataset according to the Linked Open Data principles and connect it to external datasets. Then, an integration with Open Street Map allows to display the accident data on a map. Here, the final user is able to identify which road sections are most critical based on the number of deaths, injuries or accidents.

No abstract