Inductive Link Prediction for Nodes Having Only Attribute Information

Authors: Yu Hao, Xin Cao, Yixiang Fang, Xike Xie, Sibo Wang

IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on several benchmark datasets show that our proposed model significantly outperforms existing inductive link prediction methods, and also outperforms the state-of-the-art methods on transductive link prediction.
Researcher Affiliation Academia 1University of New South Wales 2University of Science and Technology of China 3The Chinese University of Hong Kong
Pseudocode Yes Algorithm 1 The learning process of DEAL
Open Source Code No The paper does not provide an explicit statement or link for the open-sourcing of the code for the described methodology.
Open Datasets Yes CS ([Shchur et al., 2018]) 18,333 81,894 6,805 PPI ([Zitnik and Leskovec, 2017]) 1,767 16,159 50 Cora ([Mc Callum et al., 2000]) 2,708 5,278 1,433 Cite Seer ([Sen et al., 2008]) 3,327 4,552 3,703 Pub Med ([Namata et al., 2012]) 19,717 44,324 500 Computers ([Mc Auley et al., 2015]) 13,752 245,861 767 Photo ([Mc Auley et al., 2015]) 7,650 119,081 745
Dataset Splits Yes Inductive link prediction. For the inductive case, the nodes in the test set are unseen during the training process. Similar to the dataset split setting of [Bojchevski and G unnemann, 2018], we randomly hide 10% nodes and use the edges between them for the test set. The remaining nodes and edges are used for training and validation. Transductive link prediction. For the transductive case, all the nodes on the graph can be seen during the training. Similar to the dataset split setting of [You et al., 2019], we randomly sample 10%/10% edges and an equal number of non-edges as validation/test set. The remaining non-edges and 80% edges are used as the training set.
Hardware Specification No The paper does not provide specific hardware details used for running experiments.
Software Dependencies No The paper mentions general neural network components (e.g., MLP, GCN, GIN, GAT) but does not provide specific software names with version numbers for libraries or frameworks used (e.g., PyTorch, TensorFlow, scikit-learn versions).
Experiment Setup Yes In all the experiments, the default embedding size is 64. For each training mini-batch, the linked node pairs account for 40%. We tune the hyper-parameters of baseline models and our proposed DEAL with the grid search algorithm on the validation set.