Knowledge and temporal graphs: research map
1,162 papers on Knowledge and temporal graphs in Graph learning, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 3 clusters and 12 approaches. The busiest year so far is 2026.
What it is
Graphs whose nodes and edges carry meaning: knowledge graphs of facts, scene graphs of objects in an image, and graphs that change over time such as traffic networks. Models learn embeddings of entities and relations to answer queries, complete missing facts and forecast.
Terms you will see on this page:
- Knowledge graph
- facts stored as (head, relation, tail) triples.
- Link prediction
- inferring missing edges or facts.
- Temporal graph
- a graph whose edges or features change over time.
Start reading with the most cited papers here:
- Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · 2,091 citations)
- Actional-Structural Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · 1,272 citations)
Within Graph learning, its share shrank from 20.3% in 2023–24 to 18.0% in 2025–26 (258 → 283 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Knowledge and temporal graphs in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Continuous-time dynamic graph learning rising: 15% → 26% of the topic (38 → 73 papers)
- Spatio-temporal graph traffic forecasting rising: 4% → 7% of the topic (11 → 19 papers)
- Social and behavior graphs rising: 10% → 12% of the topic (26 → 35 papers)
- Graph-based dialog systems falling: 10% → 6% of the topic (27 → 16 papers)
- Knowledge graph embeddings falling: 8% → 5% of the topic (20 → 13 papers)
Papers from 2025–2026 that are already cited most:
- Spatiotemporal-aware Trend-Seasonality Decomposition Network for Traffic Flow Forecasting (AAAI 2025 · 49 citations)
- LS-TGNN: Long and Short-Term Temporal Graph Neural Network for Session-Based Recommendation (AAAI 2025 · 11 citations)
Approaches and key papers
Graphs for vision and social data · 555 papers
Keywords: scene, image, visual
Approaches in this cluster:
- Graph reasoning for visual recognition (140 papers)
Model relations among regions, actors, and skeleton joints using graph networks. - Graph-based dialog systems (102 papers)
Use graph structures for dialog reasoning and language model context. - Heterogeneous graph learning (70 papers)
Learn from and explain heterogeneous graphs with transfer and relational methods. - Social and behavior graphs (95 papers)
Model social interactions with graphs for trajectory and behavior prediction. - Scene graph generation (148 papers)
Generate scene graphs with unbiased relation learning and transformers.
Most cited and most cited since 2024:
- Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · 2,091 citations)
- Actional-Structural Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · 1,272 citations)
- BlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition (CVPR 2024 · 176 citations)
- LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024 · 146 citations)
Knowledge graph reasoning · 320 papers
Keywords: knowledge graph, entities, kg
Approaches in this cluster:
- Knowledge graph embeddings (120 papers)
Embed entities and relations with geometric transformations for link prediction. - Temporal knowledge graph reasoning (89 papers)
Forecast and reason over time-evolving knowledge graphs using rules and sequence models. - Logical query answering and rule reasoning (73 papers)
Answer logical queries and induce rules with neuro-symbolic path-based reasoning. - Inductive knowledge graph completion (38 papers)
Complete knowledge graphs for unseen entities using relational semantics and GNNs.
Most cited and most cited since 2024:
- RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (ICLR 2019 · 756 citations)
- Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning (ICLR 2018 · 398 citations)
- MKG-FENN: A Multimodal Knowledge Graph Fused End-to-End Neural Network for Accurate Drug–Drug Interaction Prediction (AAAI 2024 · 41 citations)
- MINES: Message Intercommunication for Inductive Relation Reasoning over Neighbor-Enhanced Subgraphs (AAAI 2024 · 34 citations)
Temporal and dynamic graphs · 287 papers
Keywords: dynamic, temporal graph, spatio temporal
Approaches in this cluster:
- Continuous-time dynamic graph learning (141 papers)
Model link prediction on continuous-time dynamic graphs and benchmark temporal graph learning. - Temporal graph neural architectures (77 papers)
Design state-space, motif-based, and spiking architectures for dynamic graphs. - Spatio-temporal graph traffic forecasting (69 papers)
Forecast traffic with spatio-temporal graph neural networks.
Most cited and most cited since 2024:
- Disentangling and Unifying Graph Convolutions for Skeleton-Based Action Recognition (CVPR 2020 · 1,219 citations)
- Structural-RNN: Deep Learning on Spatio-Temporal Graphs (CVPR 2016 · 1,151 citations)
- Spatio-Temporal Pivotal Graph Neural Networks for Traffic Flow Forecasting (AAAI 2024 · 129 citations)
- Temporal Graph Contrastive Learning for Sequential Recommendation (AAAI 2024 · 62 citations)
Related topics in Graph learning
- Graph algorithms and clustering (990)
- GNN architectures (1,634)
- Graph self-supervised learning (1,399)
