ClimatePolicyRadar/rag-climate-expert-eval

这是一个由气候政策雷达组织与志愿者领域专家合作创建的数据集，包含了1,138个查询的2,361个标注，针对多种不同的RAG管道。数据集主要用于评估RAG系统在气候和能源领域的表现，标注指标包括总体质量、响应情况、相关性、准确性、CPR生成策略等。数据集由来自不同国际组织的16位能源专家以个人身份标注。

This is a dataset created in collaboration with Climate Policy Radar and volunteer domain experts, containing 2,361 annotations on 1,138 queries across various RAG pipelines. The dataset is primarily used to evaluate the performance of RAG systems in the climate and energy domain, with annotation metrics including overall quality, responsiveness, relevance, fidelity, CPR generation policy, etc. The dataset has been annotated by 16 energy experts from various international organizations in a personal capacity.