Topological Preservation in Temporal Link Prediction

Marco Campos, William Ott, Henry Adams, Sarah E. Simpson, Dan J. Krofcheck, Casey Doyle, Michael Xi

Track: Full Paper (8 pages)

Keywords: Temporal, link, prediction, topological, data, analysis, zigzag, persistence, graph, neural, networks

TL;DR: We demonstrate the use of zigzag persistence to evaluate temporal topology preservation in temporal link prediction models and introduce a simplified model for enhancing interpretability in their outputs.

Abstract: Temporal link prediction seeks to model evolving networks to forecast future or missing interactions. Although many methods in this field achieve strong predictive performance, interpretability remains limited, especially in high-stakes domains. We address this by showing how topological data analysis can assess the faithfulness of learned representations to the underlying data, providing a pipeline for comparing temporal topological structure across model output. We further introduce a prototypical model that enables this analysis while maintaining predictive power. Taken together, these contributions lay the groundwork for models whose representations are more transparent to end users.

Submission Number: 4