Graph-based analysis of transportation networks

This study aims to analyze vehicle passage profiles within a city's transportation network to identify critical intersections, green wave corridors, link predictions, and maximum traffic flow. Using an original dataset derived from 54 intersections in Malatya and 147 connections between these intersections, the city's transportation network was examined in detail. Centrality analyses employed classical centrality algorithms such as PageRank, Closeness, and Betweenness, alongside the Malatya centrality method. These methods were utilized to determine the traffic intensity and strategic importance of intersections, offering new perspectives for urban traffic management. In the context of community detection, previously applied methods such as Walktrap, Optimal, and Spinglass were also included, enabling more precise identification of subregional structures and green wave corridors within the transportation network. Furthermore, based on the existing network structure, link prediction methods using Jaccard and Cosine similarity measures were employed to forecast potential future connections, providing insights into where new roads and intersections might emerge. The application of the maximum flow algorithm facilitated the analysis of vehicle passage capacity between two points and enabled the identification of bottlenecks within the network. Analyses were conducted using the R programming language and the igraph library, with the results presented both numerically and visually in the study. These findings provide a significant decision support system for improving urban traffic planning and identifying high-traffic intersections, ultimately contributing to the enhancement of the city's transportation infrastructure.