Exploring AI Applications for Optimal Business Location Selection in Smart Cities: A Literature Review

With the rapid growth of urbanization, adaptive technologies are being used to enhance infrastructure, services and the economy of smart cities. A critical aspect in smart cities' advancement is optimizing business location selection such that economic opportunities and resource efficiency are well-balanced. To identify where new endeavors might be most valuable, large scale urban data has been recognized as an important area where Artificial Intelligence (AI) can be utilized. This study reviews AI applications for finding optimal business locations in smart cities. The objective is to understand the methodologies used, factors considered and how such models can be implemented in practical business environments. The review revealed three main findings: (i) Random Forest is the most dominant machine learning algorithm used representing around 43% of the overall studies due to its robust performance while handling enormous and complicated data sets. (ii) Most papers primarily examined demographic and geographical factors with some studies incorporating sales-related and behavioral factors. (iii) The highest accuracy reported was 98.4%, achieved by the Improved Gravity Model and PCA-BP Neural Network. While many models have potential for real-world implementation, only a few studies have assessed the models in actual business scenarios. The outcomes reveal that even though accuracy obtained through AI techniques are impressive, the lack of real-world verification merits investigation. Therefore, next studies must aim at the implementation of these models to the real world to improve their applicability in smart city planning.