JamieSJS/r-oxford

该数据集包含三个主要配置项：query、corpus和qrels。qrels配置项用于存储查询与文档之间的相关性评分，包含query-id、corpus-id和score三个特征，测试集包含3555537个样本。corpus配置项用于存储文档信息，包含id、modality和image三个特征，corpus分割包含5063个样本。query配置项用于存储查询信息，包含id、modality和image三个特征，测试集包含5063个样本。

This dataset is designed for information retrieval and image retrieval tasks, containing three configurations: qrels, corpus, and query. The qrels configuration is used to evaluate the relevance between queries and the corpus, including query ID, corpus ID, and relevance score. The corpus configuration contains the ID, modality, and image data of the corpus. The query configuration contains the ID, modality, and image data of the queries, primarily used for testing the effectiveness of image retrieval.