A Survey of Link Prediction in Temporal Networks
Jiafeng Xiong, Ahmad Zareie, Rizos Sakellariou
TL;DR
This survey addresses Temporal Link Prediction (TLP) by introducing a novel taxonomy that separates the representation unit (how temporal graphs are encoded) from the inference unit (how future links are predicted). It reviews discrete-time and continuous-time dynamic graph representations and a wide spectrum of inference strategies, organizing existing work into a framework that clarifies applicability, strengths, and gaps. Key contributions include a systematic taxonomy, a comprehensive survey of representation techniques (snapshots, feature-based, and latent-variable methods) and inference approaches (transductive and inductive, including RNN- and attention-based models), and a discussion of problem variations (directed, heterogeneous, and hypergraph temporal networks) and future directions (explainability, scalability, and CT dynamics). The framework provides guidance for constructing new TLP models by mixing representation and inference units and highlights unexplored combinations with potential for improved performance and interpretability in real-world, large-scale temporal networks.
Abstract
Temporal networks have gained significant prominence in the past decade for modelling dynamic interactions within complex systems. A key challenge in this domain is Temporal Link Prediction (TLP), which aims to forecast future connections by analysing historical network structures across various applications including social network analysis. While existing surveys have addressed specific aspects of TLP, they typically lack a comprehensive framework that distinguishes between representation and inference methods. This survey bridges this gap by introducing a novel taxonomy that explicitly examines representation and inference from existing methods, providing a novel classification of approaches for TLP. We analyse how different representation techniques capture temporal and structural dynamics, examining their compatibility with various inference methods for both transductive and inductive prediction tasks. Our taxonomy not only clarifies the methodological landscape but also reveals promising unexplored combinations of existing techniques. This taxonomy provides a systematic foundation for emerging challenges in TLP, including model explainability and scalable architectures for complex temporal networks.