Iskanje po katalogu digitalne knjižnice

Opis: Criminal activities are predominantly due to males, with females exhibiting a significantly lower involvement, especially in serious offenses. This pattern extends to organized crime, where females are often perceived as less tolerant to illegal practices. However, the roles of males and females within corruption networks are less understood. Here, we analyze data from political scandals in Brazil and Spain to shed light on gender differences in corruption networks. Our findings reveal that females constitute 10% and 20% of all agents in the Brazilian and Spanish corruption networks, respectively, with these proportions remaining stable over time and across different scandal sizes. Despite this disparity in representation, centrality measures are comparable between genders, except among highly central individuals, for which males are further overrepresented. Additionally, gender has no significant impact on network resilience, whether through random dismantling or targeted attacks on the largest component. Males are more likely to be involved in multiple scandals than females, and scandals predominantly involving females are rare, though these differences are explained by a null network model in which gender is randomly assigned while maintaining gender proportions. Our results further reveal that the underrepresentation of females partially explains gender homophily in network associations, although in the Spanish network, male-to-male connections exceed expectations derived from a null model.
Ključne besede: gender disparity, corruption network, political scandal, social physics, social physics
Objavljeno v DKUM: 25.04.2025; Ogledov: 0; Prenosov: 0
Celotno besedilo (3,50 MB)
Gradivo ima več datotek! Več...

Opis: Recent advances in deep learning methods have enabled researchers to develop and apply algorithms for the analysis and modeling of complex networks. These advances have sparked a surge of interest at the interface between network science and machine learning. Despite this, the use of machine learning methods to investigate criminal networks remains surprisingly scarce. Here, we explore the potential of graph convolutional networks to learn patterns among networked criminals and to predict various properties of criminal networks. Using empirical data from political corruption, criminal police intelligence, and criminal financial networks, we develop a series of deep learning models based on the GraphSAGE framework that are able to recover missing criminal partnerships, distinguish among types of associations, predict the amount of money exchanged among criminal agents, and even anticipate partnerships and recidivism of criminals during the growth dynamics of corruption networks, all with impressive accuracy. Our deep learning models significantly outperform previous shallow learning approaches and produce high-quality embeddings for node and edge properties. Moreover, these models inherit all the advantages of the GraphSAGE framework, including the generalization to unseen nodes and scaling up to large graph structures.
Ključne besede: organized crime, complexity, crime prediction, GraphSAGE
Objavljeno v DKUM: 20.06.2024; Ogledov: 236; Prenosov: 18
Celotno besedilo (2,36 MB)
Gradivo ima več datotek! Več...