Microsoft Academic Graph

该数据集名为MAG数据库，包含了大约1.6亿篇学术出版物，每篇出版物具有标题、作者、所属机构、发表场所、学科领域、摘要、引用次数以及文献引用等属性。该数据库支持跨学科和跨领域的分类，并且不断更新，以最小化预处理工作。其规模达到了1.6亿篇出版物，任务是对学术出版物进行分层分类。

This dataset is named the MAG Database. It encompasses approximately 160 million academic publications, with each publication equipped with attributes including title, authors, affiliated institutions, publication venues, subject domains, abstracts, citation counts, and cited references. The database supports interdisciplinary and cross-domain classification, and is constantly updated to minimize preprocessing efforts. Boasting a corpus of 160 million academic publications, its core task is hierarchical classification of academic publications.

搜集汇总

背景与挑战

背景概述

Microsoft Academic Graph是一个包含科学出版物及其引用关系、作者、机构等信息的异构图，支持多种访问方式，包括Azure存储、API和MAKES服务，用于增强Bing等微软产品的学术搜索体验。