Timeline Dataset of Metro Construction in Chinese Cities (1971-2025)

Based on the subway data of the China Public Transport Operation Network Dataset (CPTOND-2025), and innovatively uses a combination of multi language model cross validation and manual hierarchical verification to systematically construct a time-series dataset of urban subway construction in China. The dataset covers the subway systems of 46 cities in China (including mainland China, Hong Kong, Macao, and Taiwan), including 992 subway lines, 17731 subway stations, and detailed opening year information for 16759 line segments. The time span is from the trial operation of Beijing Subway Line 1 in 1971 to 2025. The research uses Kimi, DeepSeek, and Tongyi Qianwen three big language model APIs to query the opening year of each site in parallel. Through the consistency judgment of the three models, 67.0% of the data is automatically passed. For 33.0% of the data with differences, the Qianfan model is used to retrieve the official opening announcement and Baidu Encyclopedia link, supplemented by manual verification and confirmation. After multiple rounds of verification and official data comparison, the overall accuracy rate reaches 98.9% (how to evaluate this). The data is stored in Shapefile format, using WGS-84 coordinate system, including key attribute fields such as station name (Chinese and English), opening year, line, three model query results, and operating company. This dataset provides important temporal data support for the research of urban rail transit development models, spatial evolution analysis, TOD planning evaluation, etc. It is of great significance for understanding the spatiotemporal evolution laws of China's urban subway network.