Analisis Perbandingan Algoritma Dijkstra, Haversine, dan Distance Matrix API pada Penentuan Jarak Sekolah di Kota Semarang

The distance between home and school often becomes an important consideration in the school selection process, as it relates to accessibility, comfort, and travel time efficiency. There are various distance calculation methods that can be used, each with its own advantages. This study aims to compare three distance calculation methods, namely Dijkstra (using road network data from OpenStreetMap), the Haversine method, and the Google Distance Matrix API. The results show that Dijkstra provides a more realistic distance estimate compared to the Haversine method, with an average difference of 1.78 km from the Google Distance Matrix API results. Meanwhile, the Haversine method has an average difference of 3.64 km. This research offers an offline solution based on the Dijkstra algorithm for school navigation in large cities. The developed system provides an efficient and independent alternative for distance estimation for zoning selection in school admissions, without reliance on an internet connection. Nevertheless, this system has not yet considered dynamic factors such as traffic conditions, and it is still limited to the Semarang City area and has not been optimized for large-scale usage scenarios.