Timezone: »
While numerous approaches have been developed to embed graphs into either Euclidean or hyperbolic spaces, they do not fully utilize the information available in graphs, or lack the flexibility to model intrinsic complex graph geometry. To utilize the strength of both Euclidean and hyperbolic geometries, we develop a novel Geometry Interaction Learning (GIL) method for graphs, a well-suited and efficient alternative for learning abundant geometric properties in graph. GIL captures a more informative internal structural features with low dimensions while maintaining conformal invariance of each space. Furthermore, our method endows each node the freedom to determine the importance of each geometry space via a flexible dual feature interaction learning and probability assembling mechanism. Promising experimental results are presented for five benchmark datasets on node classification and link prediction tasks.
Author Information
Shichao Zhu (Institute of Information Engineering, Chinese Academy of Sciences)
Shirui Pan (Monash University)
Chuan Zhou (Chinese Academy of Sciences)
Jia Wu (Macquarie University)
Yanan Cao (Institute of Information Engineering, Chinese Academy of Sciences)
Bin Wang (Xiaomi AI Lab)
More from the Same Authors
-
2021 Poster: Contrastive Graph Poisson Networks: Semi-Supervised Learning with Extremely Limited Labels »
Sheng Wan · Yibing Zhan · Liu Liu · Baosheng Yu · Shirui Pan · Chen Gong -
2020 Poster: Differentiable Neural Architecture Search in Equivalent Space with Exploration Enhancement »
Miao Zhang · Huiqi Li · Shirui Pan · Xiaojun Chang · Zongyuan Ge · Steven Su -
2020 Poster: Graph Stochastic Neural Networks for Semi-supervised Learning »
Haibo Wang · Chuan Zhou · Xin Chen · Jia Wu · Shirui Pan · Jilong Wang