Meta-Graph: Few shot Link Prediction via Meta Learning

Avishek Joey Bose, Ankit Jain, Piero Molino, William L. Hamilton

Keywords: Meta Learning, Link Prediction, Graph Representation Learning, Graph Neural Networks

TL;DR: We apply gradient based meta-learning to the graph domain and introduce a new graph specific transfer function to further bootstrap the process.

Abstract: We consider the task of few shot link prediction, where the goal is to predict missing edges across multiple graphs using only a small sample of known edges. We show that current link prediction methods are generally ill-equipped to handle this task---as they cannot effectively transfer knowledge between graphs in a multi-graph setting and are unable to effectively learn from very sparse data. To address this challenge, we introduce a new gradient-based meta learning framework, Meta-Graph, that leverages higher-order gradients along with a learned graph signature function that conditionally generates a graph neural network initialization. Using a novel set of few shot link prediction benchmarks, we show that Meta-Graph enables not only fast adaptation but also better final convergence and can effectively learn using only a small sample of true edges.

Code: https://anonymous.4open.science/r/2212328f-4954-4798-b0f2-dbd75005c9ae/

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:1912.09867/code)