irds/mmarco_id_train

`mmarco/id/train`数据集由ir-datasets包提供，主要用于文本检索任务。该数据集包含808,731条查询（queries）、532,761条相关性评估（qrels）和39,780,811对文档对（docpairs）。文档部分需要使用`irds/mmarco_id`数据集。数据集的使用示例代码展示了如何加载查询、相关性评估和文档对。该数据集的相关研究论文为《mMARCO: A Multilingual Version of MS MARCO Passage Ranking Dataset》。

The `mmarco/id/train` dataset is provided by the ir-datasets package and is primarily used for text retrieval tasks. It contains 808,731 queries, 532,761 qrels (relevance judgments), and 39,780,811 document pairs (docpairs). For the document corpus, the `irds/mmarco_id` dataset should be utilized. The sample usage code of this dataset demonstrates how to load queries, relevance judgments, and document pairs. The associated research paper for this dataset is *mMARCO: A Multilingual Version of MS MARCO Passage Ranking Dataset*.

原始信息汇总

数据集概述

数据集名称

mmarco/id/train

数据来源

• 源数据集：irds/mmarco_id

任务类别

• 文本检索

数据内容

• queries（查询）：808,731条
• qrels（相关性评估）：532,761条
• docpairs（文档对）：39,780,811对

使用方法

python from datasets import load_dataset

加载查询数据

queries = load_dataset(irds/mmarco_id_train, queries) for record in queries: record # {query_id: ..., text: ...}

加载相关性评估数据

qrels = load_dataset(irds/mmarco_id_train, qrels) for record in qrels: record # {query_id: ..., doc_id: ..., relevance: ..., iteration: ...}

加载文档对数据

docpairs = load_dataset(irds/mmarco_id_train, docpairs) for record in docpairs: record # {query_id: ..., doc_id_a: ..., doc_id_b: ...}

引用信息

@article{Bonifacio2021MMarco, title={{mMARCO}: A Multilingual Version of {MS MARCO} Passage Ranking Dataset}, author={Luiz Henrique Bonifacio and Israel Campiotti and Roberto Lotufo and Rodrigo Nogueira}, year={2021}, journal={arXiv:2108.13897} }