RecSim Simulated Environment
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/FEBR-rec/ExpertDrivenRec
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个使用RecSim开发的模拟环境，旨在评估FEBR框架在视频推荐方面的性能，特别关注多步骤的用户与推荐系统互动。该数据集包含模拟专家在视频选择、观看和评估行为生成的轨迹，这些轨迹被用于训练和评估推荐策略。该任务专注于视频推荐评估。
This dataset is a simulation environment developed with RecSim, designed to assess the performance of the FEBR framework for video recommendation, with special focus on multi-step interactions between users and recommendation systems. It encompasses trajectories generated by simulating experts' behaviors in video selection, viewing, and evaluation, which are used for training and evaluating recommendation strategies. This task centers on video recommendation evaluation.
提供机构：
RecSim
