Scholarlink

This study presents a framework to predict socioeconomic features across Paris by integrating mobile application (app) traffic data with Points of Interest (POI) data. Using the NetMob 2023 app dataset and Humanitarian Data Exchange (HDX) POI dataset, app traffic was categorized into key groups (e.g., social media, productivity), and POIs were grouped (e.g., education, business) to provide comprehensive regional insights. Two machine learning models-Logistic Regression and K-Nearest Neighbors (KNN)-were evaluated on mobile app data alone and combined app+POI data. Results show that KNN with app+POI data outperformed Logistic Regression model, by achieving 98.7% accuracy. This highlights the value of spatial and behavioral data integration. This study demonstrates the potential of app and POI data as accessible alternatives to traditional data sources in real-time socioeconomic analysis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Multi-label Classification Framework for Socioeconomic Insights Using Mobile Traffic and POI Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Farida Hazem,
  • Doaa Mahdy,
  • Mustafa Sultan,
  • Saif Rady,
  • Noha Gamal

摘要

This study presents a framework to predict socioeconomic features across Paris by integrating mobile application (app) traffic data with Points of Interest (POI) data. Using the NetMob 2023 app dataset and Humanitarian Data Exchange (HDX) POI dataset, app traffic was categorized into key groups (e.g., social media, productivity), and POIs were grouped (e.g., education, business) to provide comprehensive regional insights. Two machine learning models-Logistic Regression and K-Nearest Neighbors (KNN)-were evaluated on mobile app data alone and combined app+POI data. Results show that KNN with app+POI data outperformed Logistic Regression model, by achieving 98.7% accuracy. This highlights the value of spatial and behavioral data integration. This study demonstrates the potential of app and POI data as accessible alternatives to traditional data sources in real-time socioeconomic analysis.