GraphLand benchmark evaluates graph ML models on diverse industrial data

Researchers have introduced GraphLand, a new benchmark comprising 14 diverse graph datasets sourced from various industrial applications. This benchmark aims to address the limitations of existing graph ML evaluations, which often focus on narrow data domains and academic networks. GraphLand allows for a more comprehensive assessment of graph machine learning models, including general-purpose graph foundation models, across a wide range of graph characteristics and realistic temporal shifts. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Provides a more robust evaluation framework for graph ML models, potentially guiding future development of graph foundation models.

RANK_REASON The cluster describes a new benchmark dataset and evaluation framework for graph machine learning models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• GraphLand
• Oleg Platonov
• arXiv

• paper
• other