Not All Memories Age the Same: Autodiscovery of Adaptive Decay in Knowledge Graphs

TL;DR

This paper introduces a hierarchical, data-driven decay model for knowledge graphs that accounts for different knowledge types' temporal dynamics, improving retrieval relevance over uniform decay methods.

Contribution

It proposes a novel decay surface parameterized by velocity and volatility, learned from data without predefined taxonomies, and demonstrates its effectiveness on synthetic and real datasets.

Findings

HDBSCAN clustering perfectly recovers hierarchical parameters (ARI=1.0) in synthetic data.

Heterogeneous decay improves retrieval performance, reducing error by 18x compared to uniform decay.

Velocity-volatility clusters align with observable persistence patterns and follow the Lindy effect.

Abstract

Knowledge graphs used for retrieval treat all facts as equally current. Existing temporal approaches apply uniform decay, using a single forgetting curve regardless of knowledge type. We show this is fundamentally misspecified: different knowledge types exhibit different temporal dynamics, and the core retrieval problem is not latency or throughput but identifying what is important at query time. We propose a hierarchical framework that replaces uniform decay with a continuous decay surface parameterized by two orthogonal signals: velocity (how frequently a concept is observed) and volatility (how much the value changes between observations, measured via embedding distance). The decay surface is decomposed into three learnable levels: domain-level parameters capture universal patterns (some predicates are inherently permanent, others inherently transient), context-level parameters capture setting-dependent variation, and entity-level adaptation personalizes decay to specific subjects. All parameters emerge from data through survival analysis on observed value lifetimes, requiring no predefined taxonomies or domain expertise. We formulate edge lifetime as a survival problem where the event is value supersession (a meaningfully different value replacing the current one), distinct from mere re-observation. Experiments on synthetic temporal knowledge graphs demonstrate recovery of planted hierarchical parameters (HDBSCAN ARI = 1.0). Validation on 107 Wikipedia articles and 1,163 patient records from the Synthea clinical EHR simulator shows that velocity-volatility clusters emerge naturally, align with observable persistence patterns, and near-universally exhibit the Lindy effect (Weibull shape k < 1). Uniform decay performs 18x worse than no temporal weighting. Heterogeneous decay recovers from this, with each hierarchy level contributing measurable improvement.