nli33/beir-corpus
收藏官方服务：
资源简介：
BeIR Corpus是一个用于信息检索（IR）评估的基准数据集集合，包含多个子数据集，如arguana、climate-fever、cqadupstack系列（覆盖Android、英语、游戏等多个主题）、dbpedia-entity、fever等，涵盖科学、技术、问答、事实核查等多种领域。这些数据集通常用于训练和测试检索模型，以评估其在多样化任务上的性能。
The BeIR Corpus is a benchmark dataset collection for information retrieval (IR) evaluation, consisting of multiple sub-datasets such as arguana, climate-fever, cqadupstack series (covering topics like Android, English, gaming, etc.), dbpedia-entity, fever, and others. It spans diverse domains including science, technology, question answering, and fact-checking, and is commonly used to train and test retrieval models for assessing performance across varied tasks.
提供机构：
nli33
