SWIM

该数据集名为SWIM，是一种自适应的多层次网络商店模拟系统，能够快速模拟网络商店运营，同时应对不断变化的工作负载。其目标是最大化效用函数，通过适应不断变化的条件来实现。SWIM模拟提供了洞察服务器调整和推荐生成对用户满意度和商业指标影响的视角。该数据集能够模拟在不同工作负载下，多种服务器配置下的用户满意度、收入和成本。其任务是对Chat4Xai在深度强化学习决策制定背景下产生的解释进行忠实度和稳定性的评估。

This dataset, named SWIM, is an adaptive multi-level online store simulation system that can rapidly simulate online store operations while coping with dynamically changing workloads. Its core objective is to maximize the utility function by adapting to evolving operational conditions. The SWIM simulation provides insights into the impacts of server tuning and recommendation generation on user satisfaction and business metrics. This dataset enables the simulation of user satisfaction, revenue, and costs across various server configurations under different workloads. Its designated task is to evaluate the fidelity and stability of explanations generated by Chat4Xai in the context of deep reinforcement learning-based decision-making.