takarajordan/WorldScenario_20K
收藏Hugging Face2024-12-12 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/takarajordan/WorldScenario_20K
下载链接
链接失效反馈官方服务：
资源简介：
WorldScenario 20K是一个包含20,000个合成的多利益相关者场景的数据集，旨在模拟现实世界的决策过程。每个场景都涉及独特的环境、社会或经济问题，并包含背景信息、领域事实、利益相关者、互动对话、政策选项以及理由和评估等核心组件。数据集的主要应用包括训练AI模型、场景模拟和教育工具。数据集分为训练、验证和测试三个部分，格式为JSONL。
WorldScenario 20K is a novel dataset of 20,000 synthetically generated multi-stakeholder scenarios designed to simulate real-world decision-making processes. Each scenario explores a unique environmental, societal, or economic issue and includes core components such as background information, domain facts, stakeholders, interactive dialogue, policy options, and rationale and evaluation. The primary applications of the dataset include training AI models, scenario simulations, and educational tools. The dataset is divided into training, validation, and test splits and is formatted as JSONL.
提供机构：
takarajordan
