CHRONOS: Temporally-Aware Multi-Agent Coordination for Evolving Data Marketplaces

2026-05-22

AI Summary

This work addresses three interrelated challenges in static data markets operating over dynamic temporal knowledge graphs: outdated indexing, inaccurate Shapley-based pricing, and exhaustion of privacy budgets across multiple agents. To tackle these issues, the authors propose CHRONOS, a three-tier architecture that first models temporal decay of shortcut edges using neural ordinary differential equations (neural ODEs), then dynamically refines Shapley valuations via change-point detection, and finally coordinates multi-agent interactions through an EXP3-IX mechanism to jointly ensure differential privacy and regret control. This is the first framework to integrate neural ODEs, change-point-aware Shapley valuation, and multi-agent privacy-preserving mechanisms, providing theoretical error bounds and zero-concentrated differential privacy (zCDP) guarantees. Experiments on four benchmarks demonstrate strong performance—achieving Recall@10 of 0.937, 2.74 queries per second, and 161ms latency under ε=4.25 and δ=1e⁻⁶—while scaling efficiently to markets with up to 500 sellers.

Abstract

Temporal knowledge-graph data marketplaces face three coupled failures in static designs: stale hybrid index shortcuts reduce recall as edges evolve, stationary Shapley pricing misattributes value after distribution shifts, and uncoordinated agents over-consume a shared differential-privacy budget. We present CHRONOS, a three-layer architecture providing a unified treatment of these challenges with explicit public and private separation. Layer one applies neural-ODE temporal decay to shortcut edges, providing a per-query expected recall-loss bound of Big-O of Pq lambda delta t, with a monotone-envelope guarantee reducing bound looseness to 1.8 to 3.2 times observed loss. Layer two conditions Shapley valuation on detected changepoints and provides finite-sample error guarantees under noise. Layer three uses EXP3-IX to achieve Big-O of the square root of T log T regret while enforcing epsilon and delta differential privacy via moments accounting. CHRONOS releases a privatized affinity matrix per epoch using the Gaussian mechanism; all retrieval and ranking are post-processing, incurring no extra privacy cost. We provide multi-epoch settlement, scalability analysis for 500 sellers, and comparisons against accelerated baselines. Across four benchmarks, CHRONOS shows 0.937 recall at ten, 2.74 queries per second, 161 ms latency, and total epsilon of 4.25 at delta of 10 to the power of negative 6 under zCDP composition. These results indicate a competitive operating point. A limitation is that at this privacy level, released valuations remain noise-dominated; utility derives primarily from public index routing and adaptive scheduling driven by low-sensitivity statistics.

Problem

• temporal knowledge graph
• data marketplace
• differential privacy
• Shapley value
• multi-agent coordination

Innovation

• Temporal Knowledge Graph
• Differential Privacy
• Shapley Value
• Neural ODE
• Multi-Agent Coordination