Customer Segmentation in a Travel Agency Dataset using Clustering Algorithms
Clodomir Santana, Pedro Alexandre de Araújo Aguiar, Carmelo J. A. Bastos-Filho · 2018
In this paper, we analyze the performance of K-means, Fuzzy C-means, and Particle Swarm Clustering algorithms for clustering users with a similar profile of a travel agency. These clusters can act as target groups for advertisements designed to them, which could help to improve the advertisement effectiveness. We used 13 characteristics available in the dataset in the clustering process. The results indicate that the best number of cluster for the database is equal to 2 or 3, depending on the deployed algorithm. Moreover, all three algorithms were able to produce groups which could be used as the agency's customer segments. Regarding the metrics, some of them tend to indicate a not feasible number of clusters for the tackled problem.