This paper proposes a novel framework for targeted advertising on public screens (billboards), powered by a Raspberry Pi for efficient and cost-effective operation. It leverages data analytics to deliver personalized ads based on real-time factors like viewer demographics (gender, age) and weather conditions. Convolutional neural networks (CNNs) trained on a large, balanced dataset achieved high accuracy (95.7%) in gender prediction and promising results (85% accuracy) for age group estimation. To manage user preferences, a system is built with a database and user interface, utilizing real-time weather data from an application programming interface (API). This research demonstrates the effectiveness of data-driven advertising and the real-world potential of the smart billboard framework, offering advertisers the potential to increase engagement and revenue through highly targeted messaging.

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

Cost-Effective Real-Time Targeted Advertising on Public Screens Based on Gender, Age, and Weather Conditions

  • Hagos L. Shifare,
  • Madhu Shukla,
  • Nishant Kothari,
  • Bewketu Kehali,
  • Selam Kedre,
  • Michael Demissie

摘要

This paper proposes a novel framework for targeted advertising on public screens (billboards), powered by a Raspberry Pi for efficient and cost-effective operation. It leverages data analytics to deliver personalized ads based on real-time factors like viewer demographics (gender, age) and weather conditions. Convolutional neural networks (CNNs) trained on a large, balanced dataset achieved high accuracy (95.7%) in gender prediction and promising results (85% accuracy) for age group estimation. To manage user preferences, a system is built with a database and user interface, utilizing real-time weather data from an application programming interface (API). This research demonstrates the effectiveness of data-driven advertising and the real-world potential of the smart billboard framework, offering advertisers the potential to increase engagement and revenue through highly targeted messaging.