Guidelines for Abstracts to be presented at the China Scientific Data 面向低资源语言教育与智能分析的中小学藏文作文数据集

Spanning the period from 2011 to 2026, this dataset comprises a total of 1,187 essays. The corpus integrates content from both print publications and online resources, encompassing authoritative works—such as the *Selected Essays by Tibetan Primary School Students*—as well as texts sourced from educational websites. During the data acquisition process, Optical Character Recognition (OCR) technology was employed to digitize print texts, supplemented by manual proofreading to correct recognition errors and formatting issues. Online materials were acquired through a combined approach of targeted web crawling and manual curation, thereby ensuring the authenticity and diversity of the data. The dataset features six structured fields: essay ID, title, thematic category, educational stage, author, and body text. Regarding thematic annotation, the essays are categorized into six groups—including "Nature and Daily Life" and "Campus Life and Growth"—based on the distribution of humanistic themes found in the standardized Chinese language textbooks prescribed by the Ministry of Education.The dataset consists of one Excel (.xlsx) file, in which: (1) the file contains three worksheets corresponding to three educational stages: senior high school, junior high school, and primary school; (2) each worksheet includes five columns: Composition ID, Title, Topic Category, Educational Stage, and Full Text. Spanning the period from 2011 to 2026, this dataset comprises a total of 1,187 essays. The corpus integrates content from both print publications and online resources, encompassing authoritative works—such as the *Selected Essays by Tibetan Primary School Students*—as well as texts sourced from educational websites. During the data acquisition process, Optical Character Recognition (OCR) technology was employed to digitize print texts, supplemented by manual proofreading to correct recognition errors and formatting issues. Online materials were acquired through a combined approach of targeted web crawling and manual curation, thereby ensuring the authenticity and diversity of the data. The dataset features six structured fields: essay ID, title, thematic category, educational stage, author, and body text. Regarding thematic annotation, the essays are categorized into six groups—including "Nature and Daily Life" and "Campus Life and Growth"—based on the distribution of humanistic themes found in the standardized Chinese language textbooks prescribed by the Ministry of Education.The dataset consists of one Excel (.xlsx) file, in which: (1) the file contains three worksheets corresponding to three educational stages: senior high school, junior high school, and primary school; (2) each worksheet includes five columns: Composition ID, Title, Topic Category, Educational Stage, and Full Text.