Research Seminar by ZHANG Ce | Representation Learning on Homogeneous and Heterogeneous Graphs

In this talk, we will be sharing two papers accepted by ECML/PKDD 2021. The first paper, Representation Learning on Multi-Layered Heterogeneous Network (link), proposes intra-layer and cross-layer proximity concepts to learning node embeddings on multi-layered heterogeneous networks. Intra-layer proximity simulates propagation along homogeneous nodes to explore latent structural similarities. Cross-layer proximity captures network semantics by extending heterogeneous neighborhood across layers. Experiments showcase the advantage of our model against baselines.

The second paper, Semi-Supervised Semantic Visualization for Networked Documents (link), designs a novel semantic visualization model for homogeneous networks that incorporates partial labels. We introduce coordinate-based label distribution and label-dependent topic distribution to visualize documents, topics, and labels in a semi-supervised way. We further derive three variants for singly-labeled, multi-labeled, and hierarchically-labeled documents. Experiments verify the efficacy of our model against baselines.