Toward a Manifold-Preserving Temporal Graph Network in Hyperbolic Space

Authors: Viet Quan Le, Viet Cuong Ta

IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental By evaluating on diverse real-world dynamic graphs, our model has achieved significant improvements in link prediction and new link prediction tasks, in comparison with other baselines.
Researcher Affiliation Academia Viet Quan Le and Viet Cuong Ta Human Machine Interaction Laboratory, VNU University of Engineering and Technology, Hanoi, Vietnam {quanle9211@gmail.com, cuongtv@vnu.edu.vn}
Pseudocode Yes Algorithm 1 HMPTGN learning process
Open Source Code Yes Our implementation is available at the github repository https://github.com/quanlv9211/HMPTGN.
Open Datasets Yes We evaluate our model and other baselines on 6 datasets: email communication networks Enron [Klimt and Yang, 2004]; academic co-author networks (COLAB) [Yang and Leskovec, 2012]; private messaging network system among students (UCI) [Panzarasa et al., 2009]; synthetic dataset based on the SIR disease spreading model (Disease) [Bjørnstad et al., 2002]; interactions network on the Math Overflow website (MO) [Paranjape et al., 2016]; social network graph of Facebook Wall posts (FB) [Yang et al., 2021].
Dataset Splits No We choose the last k snapshots as the test set and the rest as the training set. The paper does not explicitly mention a separate validation set split or its details.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers like Python 3.8, CPLEX 12.4) needed to replicate the experiment.
Experiment Setup No The paper does not contain specific experimental setup details (concrete hyperparameter values, training configurations, or system-level settings) in the main text.