Learning on Graphs Conference 2026 (LoG 2026)

The Learning on Graphs Conference 2026 (LoG 2026) is a dedicated venue for research on machine learning methods that operate on graph-structured data, a rapidly growing area with applications spanning molecules, social networks, knowledge bases, transportation systems, and recommendation engines. Topics include graph neural networks, graph transformers, geometric deep learning, graph representation learning, scalable graph algorithms, knowledge graph completion, and theoretical understanding of expressive power and generalization on graphs.

LoG 2026 welcomes researchers from machine learning, data mining, theoretical computer science, chemistry, biology, and social science who work with relational and structured data. Practitioners building graph-based recommendation systems, drug discovery pipelines, or fraud detection models will find the applied tracks directly relevant. The conference has rapidly established itself as a community hub for graph ML, and graduate students entering the field will benefit greatly from its focused scope, high-quality tutorials, and accessible research culture.

graph neural networks machine learning on graphs geometric deep learning graph representation learning network analysis knowledge graphs graph algorithms graph transformers