Code and data releases from the Data Lab, grouped by research area. Every repository below is tied to a specific paper — if you are reproducing our results or building on them, start here. Everything lives under github.com/SUDataLab.
Looking for the papers themselves? See publications. For our datasets and interactive tools, see projects & tools.
Misinformation & Fake News
SAFE— Similarity-aware multi-modal fake news detection — compares what a story says against the images it carries.
Zhou, Wu, and Zafarani, PAKDD 2020HERO— Linguistic-style-aware neural networks for fake news detection.
Zhou, Li, Li, and Zafarani, arXiv:2301.02792SLIM— Fake news detection when you have very little information to work with.
Cao, Nguyen, and Zafarani, SIGKDD Explorations 2025ReCOVery— A multimodal repository for COVID-19 news credibility research: news articles, images, and their spread on Twitter.
Zhou, Mulay, Ferrara, and Zafarani, CIKM 2020CHECKED— The first Chinese COVID-19 fake news dataset, collected from Weibo.
Yang, Zhou, and Zafarani, Social Network Analysis and Mining 2021
Network Representations
KroneckerHull— Represent a whole network as a 3D convex polyhedron using stochastic Kronecker graphs — the original network-shapes method.
Jin and Zafarani, ICDM 2018SpectralPath— Represent a network as a 3D path connecting the spectral moments of the network and its subgraphs.
Jin, Tian, Li, and Zafarani, KDD 2022SpectralRobustness— A spectral measure for network robustness: assessment, design, and evolution.
Jin, Ma, Li, Eftekharnejad, and Zafarani, ICKG 2022NetworkShapesDataset— Networks and tooling behind the network-shapes work, with configurable sampling, embedding, and shape fitting.
Companion to the network shapes papers and WebShapesHON-tools— Tools for higher-order network representation and learning.
Tian and Zafarani, SIGKDD Explorations 2024common_neighbor_structure— Exploiting the common neighbor graph for link prediction.
Tian and Zafarani, CIKM 2020
Graph Sparsification & Graph Neural Networks
Graph-Sparsification-with-Graph-Convolutional-Networks— SGCN — a graph sparsifier built on graph convolutional networks.
Li, Zhang, Tian, Jin, Fardad, and Zafarani, PAKDD 2020Ultra-sparsifier— Semi-supervised graph ultra-sparsifier using reweighted L1 optimization.
Li, Zhang, Jin, and Zafarani, ICASSP 2023ADVERSPARSE— An adversarial attack framework for deep spatial-temporal graph neural networks.
Li, Zhang, Jin, Fardad, and Zafarani, ICASSP 2022
Emotions & Sentiment in Networks
- Emotions Dataset — user emotions crawled from LiveJournal, pre-processed so that every result in the paper can be reproduced. Includes friendships, followers, community memberships, and mood-annotated posts collapsed to positive, negative, or neutral polarity. User and community identifiers are anonymised.
Jin and Zafarani, CIKM 2017
Several more datasets are browsable without touching a repository, at datasets.syr.edu.