Graph link prediction in computer networks using Poisson matrix factorisation

Abstract

Graph link prediction is an important task in cyber-security: relationships between entities within a computer network, such as users interacting with computers, or system libraries and the corresponding processes that use them, can provide key insights into adversary behaviour. Poisson matrix factorisation (PMF) is a popular model for link prediction in large networks, particularly useful for its scalability. In this article, PMF is extended to include scenarios that are commonly encountered in cyber-security applications. Specifically, an extension is proposed to explicitly handle binary adjacency matrices and include known categorical covariates associated with the graph nodes. A seasonal PMF model is also presented to handle seasonal networks. To allow the methods to scale to large graphs, variational methods are discussed for performing fast inference. The results show an improved performance over the standard PMF model and other statistical network models.

References in corpus (17)

• A Comprehensive Survey on Graph Neural Networks
• Variational Inference: A Review for Statisticians
• Link Prediction in Complex Networks: A Survey
• Hierarchical structure and the prediction of missing links in networks
• Temporal Link Prediction using Matrix and Tensor Factorizations
• Link Prediction Based on Graph Neural Networks
• Latent Space Models for Dynamic Networks
• Statistical inference on random dot product graphs: a survey
• Infinite Edge Partition Models for Overlapping Community Detection and Link Prediction
• Dynamic Poisson Factorization
• Content-boosted Matrix Factorization Techniques for Recommender Systems
• Bayesian Poisson Tucker Decomposition for Learning the Structure of International Relations
• Machine Learning in Cyber-Security - Problems, Challenges and Data Sets
• On the identifiability of Bayesian factor analytic models
• Modeling item--item similarities for personalized recommendations on Yahoo! front page
• Flexible Low-Rank Statistical Modeling with Side Information
• Scalable Gaussian Process Inference with Finite-data Mean and Variance Guarantees

Cited by in corpus (2)

• Changepoint detection on a graph of time series
• Dynamically Modelling Heterogeneous Higher-Order Interactions for Malicious Behavior Detection in Event Logs