Leveraging Non-linear Dimension Reduction and Random Walk Co-occurrence for Node Embedding
This work addresses node embedding for graph analysis, but it is incremental as it builds on existing techniques with modest gains.
The authors tackled the problem of node embedding by proposing COVE, a high-dimensional embedding method that, when reduced with UMAP, slightly improves clustering and link prediction performance, performing similarly to the Louvain algorithm in community detection benchmarks.
Leveraging non-linear dimension reduction techniques, we remove the low dimension constraint from node embedding and propose COVE, an explainable high dimensional embedding that, when reduced to low dimension with UMAP, slightly increases performance on clustering and link prediction tasks. The embedding is inspired by neural embedding methods that use co-occurrence on a random walk as an indication of similarity, and is closely related to a diffusion process. Extending on recent community detection benchmarks, we find that a COVE UMAP HDBSCAN pipeline performs similarly to the popular Louvain algorithm.