Modelling sparsity, heterogeneity, reciprocity and community structure in temporal interaction data

Authors: Xenia Miscouridou, Francois Caron, Yee Whye Teh

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct experiments on real-world temporal interaction data and show that the proposed model outperforms competing approaches for link prediction, and leads to interpretable parameters. and Section 6 presents experiments on four real-world temporal interaction datasets.
Researcher Affiliation Collaboration 1Department of Statistics, University of Oxford 2Deep Mind
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes We use the MCMC scheme of (Todeschini et al., 2016) and the accompanying software package SNet OC2 to perform inference. 2https://github.com/misxenia/SNet OC
Open Datasets Yes We perform experiments on four temporal interaction datasets from the Stanford Large Network Dataset Collection3 (Leskovec and Krevl, 2014): The EMAIL dataset... The COLLEGE dataset... The MATH overflow dataset... The UBUNTU dataset... 3https://snap.stanford.edu/data/
Dataset Splits No For each dataset, we make a train-test split in time so that the training datasets contains 85% of the total temporal interactions. (Explanation: The paper describes a train-test split but does not specify a separate validation set or its details.)
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, processor types, or memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using the software package 'SNet OC2' but does not provide a specific version number for it or other software dependencies.
Experiment Setup Yes The number of communities p is set to p = 4 for the EMAIL dataset... p = 2 for the COLLEGE dataset... and p = 3 for the MATH and UBUNTU datasets... We use 100000 iterations for the first stage and 10000 for the second one.