Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining 2018
DOI: 10.1145/3219819.3219953
|View full text |Cite
|
Sign up to set email alerts
|

Interactive Paths Embedding for Semantic Proximity Search on Heterogeneous Graphs

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
39
2

Year Published

2018
2018
2021
2021

Publication Types

Select...
6
2
1

Relationship

0
9

Authors

Journals

citations
Cited by 51 publications
(41 citation statements)
references
References 23 publications
0
39
2
Order By: Relevance
“…Some recent efforts learn graph embedding models for relevance [11,12]. However, they did not show better performance than other methods in our experiments.…”
Section: Supervised Relevance Searchcontrasting
confidence: 64%
See 1 more Smart Citation
“…Some recent efforts learn graph embedding models for relevance [11,12]. However, they did not show better performance than other methods in our experiments.…”
Section: Supervised Relevance Searchcontrasting
confidence: 64%
“…To compare with the state of the art, we chose five strong baselines: PRA [9], RelSim [22], RelSUE [6], ProxE [11], and D2AGE [12]. We intended to also compare with FSPG [15], but we could not obtain its implementation from its authors and we failed to re-implement it due to some missing details in the algorithm.…”
Section: Baselinesmentioning
confidence: 99%
“…For example, friendship links between two users in a social network can be missing even they actually know each other in real world. The goal of link prediction is to infer the existence of new interactions or emerging links between users in the future, based on the observed links and the network evolution mechanism (Lü and Zhou, 2011;Al Hasan and Zaki, 2011;Liben-Nowell and Kleinberg, 2007). In network embedding, an effective model is expected to preserve both network structure and inherent dynamics of the network in the low-dimensional space.…”
Section: Link Predictionmentioning
confidence: 99%
“…Moreover, it requires the involved meta-paths to be specified as input, while our method is completely unsupervised and can automatically select aspect using statistics of the given HIN. Embedding in the context of HIN has also been studied to address various application tasks with additional supervision [3, 8, 11, 26, 27]. These methods either yield features specific to given tasks or do not generate node features, and therefore fall outside of the scope of unsupervised HIN embedding that we study.…”
Section: Related Workmentioning
confidence: 99%