Resources
Clear search
Software tools
ME2-BERT is a framework for fine-tuning a pre-trained language model like BERT to the task of moral foundation prediction. ME²-BERT integrates events and emotions for learning domain-invariant morality-relevant text representations.
SONAR is conceived to support both creators and traders in the Web3 through an interactive visualization of inspiration driven connections between NFTs, at both individual level and collection level. Demo video here.
A graph learning framework for link prediction on (attributed) multilayer networks, which originally combines (i) pairwise similarities of multilayer node embeddings learned by a GNN, and (ii) node-pair structural features learned from within-/across-layer link information based on overlapping multilayer neighborhoods. Download here.
A self-supervised graph representation learning approach based on cross-view contrastive learning conceived for multilayer heterogeneous attributed networks. Download here.
A GNN framework for representation learning and semi-supervised classification in multilayer networks with attributed entities, arbitrary number of layers and intra-/inter-layer connections between nodes. Download here.
LawNetViz, a web-based tool designed to support legal research tasks and help legal professionals as well as laymen visually exploring the article connections built upon the explicit law references detected in the article contents. Demo video here.
EMCD is a software system for finding a community structure in a multilayer network by using an ensemble-based approach. Download here.
ML-LCD is a software system for finding a community that is specific to a query node and relies on limited information about the network structure. Download here.
Datasets
The NFT-70M_transactions dataset is the largest and most up-to-date collection of Non-Fungible Tokens (NFT) transactions between 2021 and 2023 sourced from OpenSea, the leading trading platform in the Web3 ecosystem.
Our dataset of the social network of about 75,000 migrated users that was used to study the 2021 TwitterMigration phenomenon — a rapid and extensive migration of Twitter users to the decentralized social platform Mastodon.
Our dataset contains the Mastodon users and instances networks as of late 2020. It was used in several studies (APNS2021, OSNEM2022, ICWSM2022) aimed at analyzing the growth of the Fediverse and Mastodon.
This is a widely recognized Instagram dump (README) that was built up here in 2014, by Andrea T. and Roberto Interdonato, and analyzed in a paper presented at the 2014 ACM Hypertext and Social Media conference. Email us to get a copy!
Software tools from Alumni
(by Lorenzo Bloise and Carmelo Gugliotta) NetMax is a Python library that provides implementation of several algorithms for the problem of Influence Maximization in Social Networks. It also addresses the problem of Competitive Influence Maximization. It works with both signed and unsigned networks, implementing progressive, semi-progressive and non-progressive diffusion models.
(by Francesca Daniele and Marco Leto) This tool enables the construction of a knowledge graph from PDF documents, including entity and relationship extraction, advanced syntactic analysis, and the construction of graphs that can be visualized using Neo4j. The tool is customized for the topic of renewable energy, although it can be adapted to deal with any topic.
(by Edoardo Misurelli) This tool is designed to identify specific aspects within a text and assign them a sentiment polarity. These labeled text-passages are then used to fine-tune a pre-trained language model for sentiment classification.