Order-driven market simulation mimics the trader behaviors to generate order streams to support interactive studies of financial strategies. In market simulator, the multi-agent approach is commonly adopted due to its explainability. Existing multi-agent systems employ heuristic search to generate order streams, which is inefficient for large-scale simulation. Furthermore, the search-based behavior calibration often leads to inconsistent trader actions under the same general market condition, making the simulation results unstable and difficult to interpret. We propose CaliSim, the first search-free calibration approach multi-agent market simulator which achieves large-scale efficiency and behavior consistency. CaliSim uses meta-learning and devises a surrogate trading system with a consistency loss function for the reproducibility of order stream and trader behaviors. Extensive experiments in the market replay and case studies show that CaliSim achieves state-of-the-art in terms of order stream reproduction with consistent trader behavior and can capture patterns of real markets.

Measuring the quality of plot is a desirable feature for computational narrative systems.One of the notions of plot quality used in narrative theory is called tellability, which can be derived from certain structural properties, namely the types of events present and the way they are connected.These structures include not only actualized events, but also take into account virtual plans and the affective valencies of events.The present paper introduces Marie-Laure Ryan's tellability principles and suggests to computationally model them using an affective multi-agent simulation system.It discusses how such an approach implies a broader understanding of plot than commonly assumed and analysis several existing narrative systems under these considerations.Furthermore, it introduces a plot-graph formalism that allows the computational representation and analysis of the extended plot understanding.An approach to automatically generating the plot-graph is suggested in the context of the introduced multi-agent simulation system.