mteb/JaCWIRRetrievalLite
收藏Hugging Face2025-12-13 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/mteb/JaCWIRRetrievalLite
下载链接
链接失效反馈官方服务：
资源简介：
JaCWIR（Japanese Casual Web IR）是一个由来自Hatena Bookmark的问题和网页元描述组成的日语数据集。这是轻量级版本，使用5个高性能模型的硬负样本构建了缩减后的语料库（302,638个文档）。数据集主要用于文本检索和文档检索任务，是多语言性为单语（monolingual）的数据集，来源于sbintuitions/JMTEB-lite。数据集中包含corpus、qrels和queries三个配置，分别对应不同的特征和分割。
JaCWIR (Japanese Casual Web IR) is a dataset consisting of questions and webpage meta descriptions collected from Hatena Bookmark. This is the lightweight version with a reduced corpus (302,638 documents) constructed using hard negatives from 5 high-performance models. The dataset is primarily used for text retrieval and document retrieval tasks, is monolingual (Japanese), and is sourced from sbintuitions/JMTEB-lite. The dataset includes three configurations: corpus, qrels, and queries, each with distinct features and splits.
提供机构：
mteb
