XChinaCity4: A comprehensive multimodal benchmark dataset tailored for Social Media Popularity Prediction (Sample Data)

Description

XChinaCity4 is a comprehensive multimodal benchmark dataset tailored for Social Media Popularity Prediction (SMPP). Unlike existing datasets (e.g., SMPD, TPIC) that rely on static, isolated posts, XChinaCity4 focuses on contemporary social media content with rich streaming media and inter-post relationships. The dataset comprises 10,872 posts collected from the official X (formerly Twitter) accounts of four major Chinese cities—Beijing, Shanghai, Hangzhou, and Shenzhen—spanning from January 1, 2023, to June 1, 2025. It encompasses text, images, videos, and metadata. *This is a sample of the XChinaCity4. If you want to obtain the complete data in advance, please send an email to 2250501026@cnu.edu.cn.

Steps to reproduce

Data was collected via X's advanced search from the following official verified accounts: @VisitBeijingcn (Beijing) @Meetinshanghai (Shanghai) @Hangzhoufeel (Hangzhou) @szdaily1 (Shenzhen)

Institutions

• Capital Normal University

  Beijing, Beijing

Categories

Licence