CLDG applies contrastive learning to dynamic graphs by sampling multiple timespan views and pulling together the same node's representations across views, outperforming 12 baselines on 7 datasets.
Combinatorial optimization with graph convolutional networks and guided tree search,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
CLDG: Contrastive Learning on Dynamic Graphs
CLDG applies contrastive learning to dynamic graphs by sampling multiple timespan views and pulling together the same node's representations across views, outperforming 12 baselines on 7 datasets.