RecSim Simulation Environment

该数据集是一个支持与用户进行顺序交互的推荐系统构建与评估的模拟平台。它模拟了用户行为和项目结构，以评估推荐算法的有效性和性能。模拟环境包括用户模型和文档模型，其中文档属性是连续的特征，采用卡尔度量表进行衡量。该环境通过平衡用户对不同内容类型的参与度，促进了对推荐策略的评估。任务是在联邦学习的背景下评估Fedslate算法的有效性。

This dataset is a simulation platform for developing and evaluating recommender systems that support sequential interactions with users. It simulates user behaviors and item structures to assess the effectiveness and performance of recommendation algorithms. The simulation environment comprises user models and document models, wherein document attributes are continuous features measured via the Carl rating scale. This environment facilitates the evaluation of recommendation strategies by balancing user engagement across diverse content types. The core task is to evaluate the effectiveness of the Fedslate algorithm within the framework of federated learning.