COSINE: Compressive Network Embedding on Large-Scale Information Networks

There is recently a surge in approaches that learn low-dimensional embeddings of nodes networks. However, for large-scale real-world networks, it’s inefficient existing to store amounts parameters memory and update them edge by edge. With the knowledge having similar neighborhoods will be close each other embedding space, we propose COSINE (COmpresSIve Network Embedding) algorithm, which reduces footprint accelerates training process parameter sharing among nodes. applies graph partitioning algorithms networks builds dependency based on results partitioning. In this way, injects prior about high-order structural information into models, makes network more efficient effective. can applied any embedding lookup method high-quality with limited less time. We conduct experiments multi-label classification link prediction, where baselines our model have same usage. Experimental show improves up 23 percent 25 prediction. Moreover, time all representation learning methods using decreases 30 70 percent.