Stefano Tognazzi's research while affiliated with Universität Konstanz and other places

Publications (14)

Chapter
Spreading phenomena arise from simple local interaction among a large number of actors through different networks of interactions. Computational modelling and analysis of such phenomena is challenging due to the combinatorial explosion of possible network configurations. Traditional (single layer) networks are commonly used to encode the heterogene...
Article
Adversarial examples are inputs to a machine learning system that result in an incorrect output from that system. Attacks launched through this type of input can cause severe consequences: for example, in the field of image recognition, a stop signal can be misclassified as a speed limit indication. However, adversarial examples also represent the...
Preprint
Full-text available
Adversarial examples are inputs to a machine learning system that result in an incorrect output from that system. Attacks launched through this type of input can cause severe consequences: for example, in the field of image recognition, a stop signal can be misclassified as a speed limit indication.However, adversarial examples also represent the f...
Article
Full-text available
The concept of role equivalence has been applied in social network analysis for decades. Early definitions recognized two social actors as role equivalent, if they have identical relationships to the same other actors. Although this rather strong equivalence requirement has been relaxed in different ways, it is often challenging to detect interesti...
Article
Differential-algebraic equations (DAEs) are a widespread dynamical model that describes continuously evolving quantities defined with differential equations, subject to constraints expressed through algebraic relationships. As such, DAEs arise in many fields ranging from physics, chemistry, and engineering. In this paper we focus on linear DAEs, an...
Article
Full-text available
Social bots are automated accounts often involved in unethical or illegal activities. Academia has shown how these accounts evolve over time, becoming increasingly smart at hiding their true nature by disguising themselves as genuine accounts. If they evade, bots hunters adapt their solutions to find them: the cat and mouse game. Inspired by advers...
Chapter
Multi-Layer Networks (MLN) generalise the traditional, single layered networks, by allowing to simultaneously express multiple aspects of relationships in collective systems, while keeping the description intuitive and compact. As such, they are increasingly gaining popularity for modelling Collective Adaptive Systems (CAS), e.g. engineered cyber-p...
Conference Paper
Full-text available
The arm race between spambots and spambot-detectors is made of several cycles (or generations): a new wave of spambots is created (and new spam is spread), new spambot filters are derived and old spambots mutate (or evolve) to new species. Recently, with the diffusion of the adversarial learning approach, a new practice is emerging: to manipulate o...
Preprint
Full-text available
The arm race between spambots and spambot-detectors is made of several cycles (or generations): a new wave of spambots is created (and new spam is spread), new spambot filters are derived and old spambots mutate (or evolve) to new species. Recently, with the diffusion of the adversarial learning approach, a new practice is emerging: to manipulate o...
Article
Since decades, genetic algorithms have been used as an effective heuristic to solve optimization problems. However, in order to be applied, genetic algorithms may require a string-based genetic encoding of information, which severely limited their applicability when dealing with online accounts. Remarkably, a behavioral modeling technique inspired...
Chapter
One of the most distinctive features of collective adaptive systems (CAS) is the presence of many individuals which interact with each other and with the environment, giving rise to a system-level behaviour that cannot be analyzed by studying the single agents in isolation. The interaction structure among the individuals of CAS is often captured by...
Conference Paper
Full-text available
We envisage a revolutionary change in the approach to spambot detection: instead of taking countermeasures only after having collected evidence of new spambot mischiefs, in a near future techniques will be able to anticipate the ever-evolving spammers.
Conference Paper
Discovering relations between chemical reaction networks (CRNs) is a relevant problem in computational systems biology for model reduction, to explain if a given system can be seen as an abstraction of another one; and for model comparison, useful to establish an evolutionary path from simpler networks to more complex ones. This is also related to...

Citations

... However, in the current version of the Internet we are starting to understand the dire implications of delegating societally relevant tasks to algorithms that, in spite of achieving unmatched performance, are affected by several issues. Among them are biases that can prevent fair outcomes [22], the lack of transparency [23], the vulnerability to attacks and manipulations [24], [25], and the huge computational and energy requirements of complex AI models (e.g., deep learning), which limit both their affordability 7 and sustainability 8 . Each of these issues represents an open scientific challenge that will require many conjoint efforts from different scientific communities (e.g., AI and machine learning, security, ethics, and more) to be tamed and solved. ...
... Finally, current state-of-the-art techniques follow a reactive approach, coming too late in advancing new bot detection solutions, only after evidence of new evolved bots [17]. In this context, research efforts always remain one step behind the malicious evolving bot development [18]. ...
... The main disadvantage of this approach is that the constraints outlined in Definition 3.1 cannot be satisfied easily by real-world and very asymmetric MLNs as these constraints are quite strict and require precisely the same amount of neigbors from the other class. Moreover, as the experimental evaluation shows in [39], because of this reason, no equivalent layers have been found on any of the benchmark MLNs. In order to relax these strict constraints we now introduce an approximate counterpart that allows a tolerance ε used to loosen those constraints. ...
... There is a struggle between developers of algorithms for bot detection and increasingly sophisticated bots. The demonstration is that from 2017 onwards, there has been an increase in bot detectors based on adversarial approaches [19] as they are suitable for managing this type of task. Novel and more sophisticated bots do not act anymore individually, but as coordinated groups [5], forming botnets to increase their impact [20], such as the Star Wars botnet [21] or the Bursty botnet [22]. ...
... Originally designed for reducing ODEs, BDE techniques have also been applied for reducing single-layer networks [30], continuous-time Markov chains (CTMCs) [36] and differential algebraic equations [37]. In particular, in [30], a property-preserving exact model reduction algorithm for networks is shown. ...
... Despite the utility of the foregoing supervised paradigm, critiques have also arisen following real-world shifts in the increasingly sophisticated conduct of information operations [17,41]. As supervised technologies grew more adept at detecting previously encountered variants of disinformation agents, the agents deployed in new influence campaigns began to evolve and deviate from known behaviors. ...
... The interviews also revealed that all four abilities of resilience were formally developed prior to the COVID-19 outbreak. The analysis however also showed that the favourable performance was obtained through many actions undertaken reactively rather than proactively, as it is instead advisable for a number of cybersecurity issues [106]. As such, [104] leaves open the question as to whether the four potentials should be developed beforehand, in order to perform resiliently during crises. ...
... Examples of applications in the field of reaction kinetics can be found in Refs. [7][8][9][10][11][12][13][14][15][16]. Especially Lapene et al. [16] give a very good overview of applications and methods in chemical physics. ...