An Algorithmic Framework for Clustering Cab Pickup Geo-points for Cab Recommender System (CRS)

The cab recommender system (CRS) has become a crucial aspect of the transportation industry, as it helps to improve the overall efficiency of the cab service. The CRS relies on the effective clustering of cab pickup geo-points to provide effective recommendations to the users. In this research paper, we propose an algorithmic framework for clustering geo-points based on their geographical proximity and density. The proposed algorithmic framework is evaluated over three different datasets. The Silhouette coefficient and Calinski-Harabasz score parameters are used to compare the outcomes of various clustering techniques. Therefore, the research paper has four objectives: Firstly, to identify the different techniques suitable for the identified datasets. Secondly, to create an analytic framework that clusters geo-points of the cab pickup. Thirdly, to evaluate the proposed algorithmic framework over different datasets using parameters like Silhouette coefficient and Calinski-Harabasz score. Finally, the paper wraps up and evaluates the outcomes of the proposed algorithmic framework to determine the best methods for clustering geo-points.