We propose $\textit{weighted inner product similarity}$ (WIPS) for neural network-based graph embedding. In addition to the parameters of neural networks, we optimize the weights of the inner product by allowing positive and negative values. Despite its simplicity, WIPS can approximate arbitrary general similarities including positive definite, conditionally positive definite, and indefinite kernels. WIPS is free from similarity model selection, since it can learn any similarity models such as cosine similarity, negative Poincaré distance and negative Wasserstein distance. Our experiments show that the proposed method can learn high-quality distributed representations of nodes from real datasets, leading to an accurate approximation of similarities as well as high performance in inductive tasks.
Structural-Aware Sentence Similarity with Recursive Optimal Transport
Research Institute of Information Technology(信息技术研究所);Beijing National Research Center for Information Science and Technology (BNRist)(北京信息科学与技术国家研究中心);Tsinghua University(清华大学)
Overlap-aware meta-learning attention to enhance hypergraph neural networks for node classification
Department of Mathematics, College of Sciences, Shanghai University(上海大学数学系);Shanghai Institute of Applied Mathematics and Mechanics, Shanghai University(上海大学应用数学与力学研究所);School of Mechanics and Engineering Science, Shanghai University(上海大学力学与工程科学学院);BNRist, KLISS, School of Software, Tsinghua University(清华大学软件学院);THUIBCS, Tsinghua University(清华大学);Qian Weichang College, Shanghai University(上海大学钱伟长学院)