2026-10-13T07:20:48Z https://www.scidb.cn/oai

10.57760/sciencedb.j00104.00155 2026-09-10T10:52:22Z

2026-09-10 MTESum doi:10.57760/sciencedb.j00104.00155 en The first publicly available Mongolian single document summary dataset MTESum. This dataset was crawled from multiple news websites, and after normalization cleaning and BPE sub word segmentation processing, combined with an improved TextRank graph model extraction method and title guided Jaccard similarity redundancy elimination strategy, a dataset of 1000 Mongolian abstracts was constructed. MTESum dataset includes news content and its corresponding summaries.  Automatic text summarization; Data set construction; Mongolian; Extractive summarization Qiqirilige Siqintu Wang Siriguleng Han Yongshun PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank