New TRicci framework sparsifies temporal graphs for efficient learning

Researchers have developed a new framework called TRicci to address the computational challenges of learning from large, dense, and rapidly changing temporal graphs. This method extends Forman-Ricci curvature to directed weighted temporal graphs, considering structural support, temporal recency, and local interaction competition. Experiments on transaction networks and benchmark datasets show that TRicci can sparsify temporal graphs by approximately 80% while reducing training and inference time by over 55%, without significantly impacting predictive performance. AI

IMPACT This new method could enable more efficient analysis of complex, evolving systems like financial networks and social platforms.

RANK_REASON Academic paper on a novel method for graph learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →