Multi-Scale attributed node embedding
Journal of Complex Networks · 2021-05-05 · 9 (2)
DOI: 10.1093/comnet/cnab014 ↗At a glance
Journal article in Journal of Complex Networks, published online on 2021-05-07. Written by Benedek Rozemberczki and 2 co-authors. Cited 506 times so far (Crossref); 42 references. The full text is freely available (open access, CC BY 4.0).
Abstract (original language)
We present network embedding algorithms that capture information about a node from the local distribution over node attributes around it, as observed over random walks following an approach similar to Skip-gram. Observations from neighbourhoods of different sizes are either pooled (AE) or encoded distinctly in a multi-scale approach (MUSAE). Capturing attribute-neighbourhood relationships over multiple scales is useful for a range of applications, including latent feature identification across disconnected networks with similar features. We prove theoretically that matrices of node-feature pointwise mutual information are implicitly factorized by the embeddings. Experiments show that our algorithms are computationally efficient and outperform comparable models on social networks and web graphs.
Cite
APA
Rozemberczki, B., Allen, C., & Sarkar, R. (2021). Multi-Scale attributed node embedding. Journal of Complex Networks, 9(2). https://doi.org/10.1093/comnet/cnab014
Vancouver
Rozemberczki B, Allen C, Sarkar R. Multi-Scale attributed node embedding. Journal of Complex Networks. 2021;9(2). doi:10.1093/comnet/cnab014
BibTeX
@article{Rozemberczki2021,
author = {Rozemberczki, Benedek and Allen, Carl and Sarkar, Rik},
title = {Multi-Scale attributed node embedding},
journal = {Journal of Complex Networks},
year = {2021},
volume = {9},
number = {2},
doi = {10.1093/comnet/cnab014}
}
EUVIMED