NETINFOF framework: Measuring and exploiting network usable information

Given a node-attributed graph, and a graph task (link prediction or node classification), can we tell if a graph neural network (GNN) will perform well? More specifically, do the graph structure and the node features carry enough usable information for the task? Our goals are (1) to develop a fast tool to measure how much information is in the graph structure and in the node features, and (2) to exploit the information to solve the task, if there is enough. We propose NETINFOF, a framework including NETINFOF PROBE and NETINFOF ACT, for the measurement and the exploitation of network usable information (NUI), respectively. Given a graph data, NETINFOF PROBE measures NUI without any model training, and NETINFOF ACT solves link prediction and node classification, while two modules share the same backbone. In summary, NETINFOF has following notable advantages: (a) General, handling both link prediction and node classification; (b) Principled, with theoretical guarantee and closed-form solution; (c) Effective, thanks to the proposed adjustment to node similarity; (d) Scalable, scaling linearly with the input size. In our carefully designed synthetic datasets, NETINFOF correctly identifies the ground truth of NUI and is the only method being robust to all graph scenarios. Applied on real-world datasets, NETINFOF wins in 11 out of 12 times on link prediction compared to general GNN baselines.