Concept Taxonomy

The 300 concepts in this learning graph are organized into 15 categories. No category exceeds 30% of total concepts.

TaxonomyID	Category Name	Description	Count
PREREQ	Prerequisites	Linear algebra, probability, calculus, PyTorch basics required before graph content	20
FOUND	Graph Fundamentals	Core graph theory: nodes, edges, properties, random graph models, network science	37
ALGO	Classical Graph Algorithms	PageRank, centrality, community detection, graph kernels, influence propagation	33
EMB	Node Embeddings	Shallow embedding methods: DeepWalk, node2vec, matrix factorization, skip-gram	17
GNN	GNN Architecture	GNN model components and variants: GCN, GraphSAGE, GAT, GIN, pooling, tasks	44
THEORY	GNN Theory	Expressiveness, WL test, over-smoothing, over-squashing, equivariance, identity	17
TRANS	Graph Transformers	Attention-based graph architectures and positional/structural encodings	12
KG	Knowledge Graphs	KG embeddings (TransE, RotatE), multi-hop reasoning, inductive KG methods	25
HETERO	Heterogeneous Graphs	Multi-relational and typed graphs: R-GCN, HGT, HAN, meta-paths	8
APP	Applications	Domain applications: recommenders, drug discovery, temporal, subgraph mining	23
SCALE	Scalability	Sampling strategies, mini-batching, graph partitioning for large-scale training	10
GEN	Generative Models	Graph generation: GraphRNN, GCPN, VGAE, DiGress, evaluation metrics	7
ADV	Advanced Topics	LLMs+GNNs, self-supervised learning, foundation models, agents on graphs	18
TRAIN	Training & Optimization	Loss functions, regularization, augmentation, normalization, data splits	20
TOOLS	Tools & Frameworks	PyTorch Geometric, NetworkX, OGB, DGL, PyKEEN, RelBench	9