Structured Neural Marked Point Processes for Interpretable Event Interaction Modeling

TL;DR

This paper introduces SNMPP, a structured neural point process model that uncovers interpretable event relationships and improves prediction in multi-class event streams.

Contribution

It proposes a novel product-form neural influence kernel enabling explicit discovery of event interactions while maintaining high modeling flexibility.

Findings

Successfully uncovers structured event relationships in synthetic and real data.

Achieves strong predictive performance compared to existing models.

Enables explicit characterization of influence topology including excitation and inhibition.

Abstract

Multi-class event streams arise in numerous real-world applications, where uncovering structured, interpretable inter-event relationships, together with accurate prediction, remains a central challenge. Existing neural point process models are highly expressive but encode event interactions in a black-box manner, preventing explicit discovery of structured dependencies. In this paper, we propose a structured neural marked point process (SNMPP) that achieves high modeling flexibility while enabling explicit event-wise and class-wise relationship discovery from data. Our model constructs a product-form neural influence kernel composed of a signed interaction network over event types and a delay-aware monotonic temporal network. This design enables explicit characterization of inter-class influence topology -- including excitation, inhibition, and neutrality -- while flexibly capturing diverse temporal decay patterns and potential influence delays. For efficient learning, we develop a stratified Monte Carlo estimator for stochastic training. Extensive experiments on synthetic and real-world benchmark datasets validate the ability of our approach to uncover structured relationships and deliver strong predictive performance.