SMART AND DYNAMIC ADVERTISING ON A CARRIER MOUNTED CUSTOMISED loT DEVICE USING GIS/SDSS AND MACHINE LEARNING

Abstract

In the past, there have been issues with advertising such as in exposure to interested customers, static location advertisement, environmental pollution and wastage of resources. In order to combat these issues, a solution of creating a customized device that offers DD-OOH (Digital Dynamic Out-Of-Home) advertising and can be mounted on a carrier such as a vehicle was sought. Data were collected from residents of Islamabad via questionnaires and google forms. This data contained questions such as the preferences of those people when it came to purchasing products. There were 13 categories of products which were: 1) fashion and beauty 2) mobile accessories 3) electronics and home appliances 4) bikes and vehicles 5) business and industrial 6) jobs 7) animals 8) books 9) sports 10) kids' items 11) health/medical 12) food and 13) agriculture. Data was also collected concerning how much they can afford to pay for a single product. This data will be collected throughout the project and is stored in our database which is stored on our server. GPS and Raspberry pi will be the main components connected to this device. Machine Leaming and GIS libraries will be used to program the Raspberry pi such that it reads the GPS coordinates and changes the advertisements on the screening device on the basis of the preferences and affordability of the customers in that area. The machine learning algorithm makes use of survey-based data in the database as training data. Clients can register online to avail of our services through an online web interface. The advertisers will pay to play their advertisement on the screen and then the riders will be paid for taking this screen to different areas of Islamabad and advertising the products on the screen, serving as an additional source of livelihood for them.