Development of a customer feedback-enhanced ATM locator using Dijkstra's algorithm and the Haversine formula

Abstract

Automated teller machines (ATMs) allow customers to perform financial transactions without visiting a bank branch. ATM locators help users find the nearest ATMs based on their location. Customer satisfaction, heavily influenced by service quality, is crucial for banks. Hence, banks should actively seek customer feedback to improve services. Existing ATM locators lack features for communicating customer feedback to banks. This study aimed to enhance the ATM location experience with an Android-based ATM Locator application. The application was developed using C# for the server backend and Java for the mobile app. It allows users to find ATMs, get detailed information, and provide feedback sent via email to participating banks. The app's effectiveness was measured using accuracy (precision & recall), user satisfaction, and response time. User feedback improved the system's strength. The application was tested by 200 students from Ogun State Polytechnic of Health and Allied Sciences. Analysis of the results showed that 81% of recommended ATM locations were relevant, and 84% of these were recognized by users. The average service rating was 4 out of 5, indicating positive feedback. However, some reviews pointed out issues with cash replenishment processes, highlighting areas for further improvement.