Statements (56)
Predicate | Object |
---|---|
gptkbp:instance_of |
gptkb:Model
gptkb:neural_networks |
gptkbp:applies_to |
Graph data
|
gptkbp:developed_by |
gptkb:Google_Research
|
gptkbp:has_applications_in |
Chemoinformatics
Biological networks Knowledge graphs |
gptkbp:has_feature |
Attention mechanism
Fine-tuning Pre-training |
gptkbp:has_limitations |
Scalability issues
Computational cost Dependency on labeled data |
gptkbp:has_research_focus |
Transfer learning
Graph embeddings Domain adaptation Multi-task learning |
https://www.w3.org/2000/01/rdf-schema#label |
Graph-BERT
|
gptkbp:improves |
Graph representation learning
|
gptkbp:is_adopted_by |
gptkb:Industry
Academia Research institutions Startups |
gptkbp:is_based_on |
gptkb:BERT
|
gptkbp:is_compared_to |
gptkb:GCN
gptkb:Cheb_Net gptkb:Fast_GCN gptkb:Graph_SAGE GAT |
gptkbp:is_documented_in |
Research papers
Conference proceedings Theses Technical reports |
gptkbp:is_evaluated_by |
Citeseer dataset
Cora dataset Pubmed dataset Reddit dataset Amazon dataset |
gptkbp:is_influenced_by |
gptkb:BERT
GNNs |
gptkbp:is_part_of |
Deep learning models
Graph-based models |
gptkbp:is_related_to |
Link prediction
Node classification Graph classification |
gptkbp:is_supported_by |
gptkb:Tensor_Flow
gptkb:Py_Torch |
gptkbp:is_used_in |
gptkb:Natural_Language_Processing
Recommendation Systems Social Network Analysis |
gptkbp:performance |
Accuracy
F1 score AUC |
gptkbp:uses |
Transformer architecture
|
gptkbp:bfsParent |
gptkb:Super_GLUE_benchmark
|
gptkbp:bfsLayer |
5
|