Abstract:Network data in real-world tends to be error-prone due to incomplete sampling, imperfect measurements, etc.; this in turn results in inaccurate results when performing network analysis or modeling, such as node classification and link prediction, on these flawed networks. In this paper, we aim to reconstruct a reliable network from a flawed, undirected, unweighted network, a process referred to network enhancement. More specifically, network enhancement aims to detect the noisy links that are observed in the network but should not exist in the real world, as well as to predict the missing links that do indeed exist in the real world yet remain unobserved. While some attempts have been made to detect either noisy links or missing links, few of these works have considered unifying these two tasks, even though they are inter-dependent and capable of mutually boosting each others’ performance. In this paper, we therefore propose E-Net, an end-to-end graph neural network model, to leverage the mutual influence of these two tasks in order to achieve both goals more effectively. On one hand, detecting noisy links can benefit the performance of missing link prediction, while on the other hand, predicting missing links can provide indirect supervision for detecting noisy link detection when the labels of these noisy links are unavailable. Moreover, by proposing a subgraph extraction mechanism based on random walk with restart, the model can be scaled up to large networks and is able to preserve the local and global structural characteristics. The experimental results on several types of large networks demonstrate that the proposed model obtains an improvement of 10.7 percent on average in terms of F1 for predicting missing links, along with an average of 3.7 percent improvement in terms of precision for detecting noisy links compared with the state-of-the-art baselines.

Robust Network Enhancement from Flawed Networks.

Enhancing Robustness of Link Prediction for Noisy Complex Networks

RobustECD: Robust Enhancement for Community Detection in Complex Networks

Measuring the Robustness of Link Prediction Algorithms under Noisy Environment

NetRL: Task-aware Network Denoising via Deep Reinforcement Learning

RobustECD: Enhancement of Network Structure for Robust Community Detection

Network Enhancement: a general method to denoise weighted biological networks

Enhancing Predictive Expert Method for Link Prediction in Heterogeneous Information Social Networks

Machine learning of network inference enhancement from noisy measurements

Network enhancement as a general method to denoise weighted biological networks

Network Embedding with Enhanced Feature Representations for Link Prediction.

Improving Link Prediction in Complex Networks by Adaptively Exploiting Multiple Structural Features of Networks

An extended self-representation model of complex networks for link prediction

Improved Graph Autoencoder for Network Reconstruction

CNN-based Prediction of Network Robustness With Missing Edges

Adaptive Similarity Function with Structural Features of Network Embedding for Missing Link Prediction

Robust and Efficient Network Reconstruction in Complex System via Adaptive Signal Lasso

RelpNet: Relation-based Link Prediction Neural Network

Robust graph regularization nonnegative matrix factorization for link prediction in attributed networks

Unifying Node Labels, Features, and Distances for Deep Network Completion.

FERN: Leveraging Graph Attention Networks for Failure Evaluation and Robust Network Design