Matches in SemOpenAlex for { <https://semopenalex.org/work/W3198791937> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3198791937 abstract "Recently, heterogeneous Graph Neural Networks (GNNs) have become a de facto model for analyzing HGs, while most of them rely on a relative large number of labeled data. In this work, we investigate Contrastive Learning (CL), a key component in self-supervised approaches, on HGs to alleviate the label scarcity problem. We first generate multiple semantic views according to metapaths and network schemas. Then, by pushing node embeddings corresponding to different semantic views close to each other (positives) and pulling other embeddings apart (negatives), one can obtain informative representations without human annotations. However, this CL approach ignores the relative hardness of negative samples, which may lead to suboptimal performance. Considering the complex graph structure and the smoothing nature of GNNs, we propose a structure-aware hard negative mining scheme that measures hardness by structural characteristics for HGs. By synthesizing more negative nodes, we give larger weights to harder negatives with limited computational overhead to further boost the performance. Empirical studies on three real-world datasets show the effectiveness of our proposed method. The proposed method consistently outperforms existing state-of-the-art methods and notably, even surpasses several supervised counterparts." @default.
- W3198791937 created "2021-09-13" @default.
- W3198791937 creator A5006897094 @default.
- W3198791937 creator A5015101640 @default.
- W3198791937 creator A5033157677 @default.
- W3198791937 creator A5047523245 @default.
- W3198791937 creator A5048485876 @default.
- W3198791937 creator A5088739065 @default.
- W3198791937 date "2021-08-31" @default.
- W3198791937 modified "2023-09-26" @default.
- W3198791937 title "Structure-Aware Hard Negative Mining for Heterogeneous Graph Contrastive Learning" @default.
- W3198791937 cites W1533861849 @default.
- W3198791937 cites W1854214752 @default.
- W3198791937 cites W2069153192 @default.
- W3198791937 cites W2095705004 @default.
- W3198791937 cites W2109480754 @default.
- W3198791937 cites W2546547051 @default.
- W3198791937 cites W2743104969 @default.
- W3198791937 cites W2809435521 @default.
- W3198791937 cites W2842511635 @default.
- W3198791937 cites W2911286998 @default.
- W3198791937 cites W2963403868 @default.
- W3198791937 cites W2963707260 @default.
- W3198791937 cites W2963782635 @default.
- W3198791937 cites W2963858333 @default.
- W3198791937 cites W2964015378 @default.
- W3198791937 cites W2964121744 @default.
- W3198791937 cites W2970971581 @default.
- W3198791937 cites W2995607862 @default.
- W3198791937 cites W3004507689 @default.
- W3198791937 cites W3007332492 @default.
- W3198791937 cites W3033039844 @default.
- W3198791937 cites W3034978746 @default.
- W3198791937 cites W3035058308 @default.
- W3198791937 cites W3080555959 @default.
- W3198791937 cites W3081963674 @default.
- W3198791937 cites W3093274308 @default.
- W3198791937 cites W3095121901 @default.
- W3198791937 cites W3095746859 @default.
- W3198791937 cites W3099206234 @default.
- W3198791937 cites W3101821705 @default.
- W3198791937 cites W3102419180 @default.
- W3198791937 cites W3104097132 @default.
- W3198791937 cites W3104717349 @default.
- W3198791937 cites W3106428938 @default.
- W3198791937 cites W3107668149 @default.
- W3198791937 cites W3109894131 @default.
- W3198791937 cites W3114303065 @default.
- W3198791937 cites W3125808012 @default.
- W3198791937 doi "https://doi.org/10.48550/arxiv.2108.13886" @default.
- W3198791937 hasPublicationYear "2021" @default.
- W3198791937 type Work @default.
- W3198791937 sameAs 3198791937 @default.
- W3198791937 citedByCount "0" @default.
- W3198791937 crossrefType "posted-content" @default.
- W3198791937 hasAuthorship W3198791937A5006897094 @default.
- W3198791937 hasAuthorship W3198791937A5015101640 @default.
- W3198791937 hasAuthorship W3198791937A5033157677 @default.
- W3198791937 hasAuthorship W3198791937A5047523245 @default.
- W3198791937 hasAuthorship W3198791937A5048485876 @default.
- W3198791937 hasAuthorship W3198791937A5088739065 @default.
- W3198791937 hasBestOaLocation W31987919371 @default.
- W3198791937 hasConcept C119857082 @default.
- W3198791937 hasConcept C124101348 @default.
- W3198791937 hasConcept C132525143 @default.
- W3198791937 hasConcept C154945302 @default.
- W3198791937 hasConcept C31972630 @default.
- W3198791937 hasConcept C3770464 @default.
- W3198791937 hasConcept C41008148 @default.
- W3198791937 hasConcept C64869954 @default.
- W3198791937 hasConcept C80444323 @default.
- W3198791937 hasConceptScore W3198791937C119857082 @default.
- W3198791937 hasConceptScore W3198791937C124101348 @default.
- W3198791937 hasConceptScore W3198791937C132525143 @default.
- W3198791937 hasConceptScore W3198791937C154945302 @default.
- W3198791937 hasConceptScore W3198791937C31972630 @default.
- W3198791937 hasConceptScore W3198791937C3770464 @default.
- W3198791937 hasConceptScore W3198791937C41008148 @default.
- W3198791937 hasConceptScore W3198791937C64869954 @default.
- W3198791937 hasConceptScore W3198791937C80444323 @default.
- W3198791937 hasLocation W31987919371 @default.
- W3198791937 hasOpenAccess W3198791937 @default.
- W3198791937 hasPrimaryLocation W31987919371 @default.
- W3198791937 hasRelatedWork W2961085424 @default.
- W3198791937 hasRelatedWork W3117549273 @default.
- W3198791937 hasRelatedWork W3168493052 @default.
- W3198791937 hasRelatedWork W4285260836 @default.
- W3198791937 hasRelatedWork W4286629047 @default.
- W3198791937 hasRelatedWork W4306321456 @default.
- W3198791937 hasRelatedWork W4306674287 @default.
- W3198791937 hasRelatedWork W4309795810 @default.
- W3198791937 hasRelatedWork W4319453009 @default.
- W3198791937 hasRelatedWork W4224009465 @default.
- W3198791937 isParatext "false" @default.
- W3198791937 isRetracted "false" @default.
- W3198791937 magId "3198791937" @default.
- W3198791937 workType "article" @default.