Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence 2019
DOI: 10.24963/ijcai.2019/594
|View full text |Cite
|
Sign up to set email alerts
|

ProNE: Fast and Scalable Network Representation Learning

Abstract: Recent advances in network embedding has revolutionized the field of graph and network mining. However, (pre-)training embeddings for very large-scale networks is computationally challenging for most existing methods. In this work, we present ProNE---a fast, scalable, and effective model, whose single-thread version is 10--400x faster than efficient network embedding benchmarks with 20 threads, including LINE, DeepWalk, node2vec, GraRep, and HOPE. As a concrete example, the single-version ProNE requires only 2… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
148
0

Year Published

2019
2019
2023
2023

Publication Types

Select...
5
3
1
1

Relationship

0
10

Authors

Journals

citations
Cited by 150 publications
(148 citation statements)
references
References 17 publications
0
148
0
Order By: Relevance
“…Experiments show that this method can effectively improve the computational efficiency of the algorithm while retaining the spectrum information of the network. ProNE [89] further uses spectral propagation to enhance network embedding on the basis of sparse matrix decomposition, so that the learned embedding can not only capture the local structure information of the network, but also obtain the global network characteristics.…”
Section: ) Knowledge Graph Completion Methods Based On Network Represmentioning
confidence: 99%
“…Experiments show that this method can effectively improve the computational efficiency of the algorithm while retaining the spectrum information of the network. ProNE [89] further uses spectral propagation to enhance network embedding on the basis of sparse matrix decomposition, so that the learned embedding can not only capture the local structure information of the network, but also obtain the global network characteristics.…”
Section: ) Knowledge Graph Completion Methods Based On Network Represmentioning
confidence: 99%
“…Ribeiro et al [32] proposed the struc2vec algorithm in 2017, which fully considered the structural similarity. Zhang et al [33] proposed the ProNE algorithm in 2019. The ProNE algorithm decomposed sparse matrix to generate fast graph embedding representation and used spectral propagation as a lifting method for graph embedding.…”
Section: Grpah Representation Learningmentioning
confidence: 99%
“…Figure 3 shows the experimental results of link prediction in the four datasets. We used DeepWalk [6], Node2vec [3], LINE [9], Spectral [12], which are popular algorithms in research and engineering to optimize. The corresponding optimization algorithm is named Mo-DeepWalk, Mo-Node2vec, Mo-LINE, Mo-Spectral.…”
Section: Evaluation Metricsmentioning
confidence: 99%