Michele Marchetti

• Marche Polytechnic University

In recent years, Transformers have revolutionized the management of Natural Language Processing tasks, and Vision Transformers (ViTs) promise to do the same for Computer Vision ones. However, the adoption of ViTs is hampered by their computational cost. Indeed, given an image divided into patches, it is necessary to compute for each layer the atten...

Threads is a new social network that was launched by Meta in July 2023 and conceived as a direct alternative to X. It is a unique case study in the social network landscape, as it is content-based like X, but has an Instagram-based growth model, which makes it significantly different from X. As it was launched recently, studies on Threads are still...

In recent years, polarization on social media has risen significantly. Social platforms often feature a range of topics that give rise to communities of users with diametrically opposed views, who tend to avoid engaging with others having different perspectives. We call these types of communities “diverging communities”. Examples include communitie...

In this paper, we propose Multilayer network-based Visual Interpreter (MuVI), a framework for visual interpretability of Convolutional Neural Networks (CNNs) based on their mapping into multilayer networks. The peculiarity of MuVI is that it constructs a pixel-level heatmap of the salient parts of an image processed by a CNN, where the importance o...

In this article, we propose a framework that uses a multimodal multilayer network to build, manage and compare User Histories in X. A User History not only considers the contents of interest to the user, as is generally the case with a User Profile. In fact, it also records the set of interactions she has made with contents, the timestamps when the...

In recent literature, multilayer networks are increasingly being used to model and manage complex scenarios. One of them well suited to be modeled and managed using multilayer networks is represented by multimodal social networks, in which nodes and edges can be of different types. In fact, in this case, each layer of the multilayer network can be...

In this paper, we propose a framework that uses the theory and techniques of (Social) Network Analysis to investigate the learned representations of a Graph Neural Network (GNN, for short). Our framework receives a graph as input and passes it to the GNN to be investigated, which returns suitable node embeddings. These are used to derive insights o...

In recent years, the huge growth in the number and variety of blockchains has prompted researchers to investigate the cross-blockchain scenario. In this setting, multiple blockchains coexist, and wallets can exchange data and money from one blockchain to another. The effective and efficient management of a cross-blockchain ecosystem is an open prob...

Wash trading is considered a highly inopportune and illegal behavior in regulated markets. Instead, it is practiced in unregulated markets, such as cryptocurrency or NFT (Non-Fungible Tokens) markets. Regarding the latter, in the past many researchers have been interested in this phenomenon from an “ex-ante” perspective, aiming to identify and clas...

In recent years different types of Residual Neural Networks (ResNets, for short) have been introduced to improve the performance of deep Convolutional Neural Networks. To cope with the possible redundancy of the layer structure of ResNets and to use them on devices with limited computational capabilities, several tools for exploring and compressing...

The concept of scope was introduced in Social Network Analysis to assess the authoritativeness and convincing ability of a user toward other users on one or more social platforms. It has been studied in the past in some specific contexts, for example to assess the ability of a user to spread information on Twitter. In this paper, we propose a new i...

Colors characterize each object around us. For this reason, the study of colors has played a key role in Artificial Intelligence (think, for instance, of image classification, object recognition and segmentation). However, there are some topics about colors still little explored. One of them concerns fabric colors. This is a particular topic since...

Information diffusion in social networks is a classic and, at the same time, very current problem. In fact, information diffusers are always looking for new techniques to disseminate information of their interest by creating backbones among them. In this paper, we focus on a specific, but very current and relevant, scenario regarding this way of pr...

This paper falls in the context of the interpretability of the internal structure of deep learning systems. In particular, we propose an approach to map a Convolutional Neural Network (CNN) into a multilayer network. Next, to show how such a mapping helps to better understand the CNN, we propose a technique for comprising it. This technique detects...

The analysis of people's comments in social platforms is a widely investigated topic because comments are the place where people show their spontaneity most clearly. In this paper, we present a network-based data structure and a related approach to represent and manage the underlying semantics of a set of comments. Our approach is based on the extr...

The analysis of people’s comments in social platforms is a widely investigated topic because comments are the place where people show their spontaneity most clearly. In this article, we present a network-based data structure and a related approach to represent and manage the underlying semantics of a set of comments. Our approach is based on the ex...

The present paper deals with nonlinear, non-monotonic data regression. This paper introduces an efficient algorithm to perform data transformation from non-monotonic to monotonic to be paired with a statistical bivariate regression method. The proposed algorithm is applied to a number of synthetic and real-world non-monotonic data sets to test its...