Abstract

Market segmentation is an important tool, fordriving an organization to achieve its goals. This studyproposes a market segmentation technique with the bindingof unsupervised and supervised learning techniques. Themethod aims to cluster international tourists who arrived inThailand for business proposes, and to classify businesstourists by using the products of an unsupervised learningtechnique as class labels. A Self-Organizing Map (SOM), KMeansand Hierarchical clustering were applied to find thebest quality of segmentation guided by the computation ofthe Silhouette index. Segment labels were used to supervisethe learning part as class labels. Multilayer Perceptron(MLP), J48 decision tree, Decision Table, OneR and NaïveBayes classifiers were used to classify the business touristdata set, and the best performance technique was preferred.The experimental results designated that K-Meansoutperformed the other clustering techniques and providedfive different segments. Moreover, the Naïve Bayes classifiergave the best performance among the other classifiers basedon the business tourist variables. Thus, this model can beused to predict the segment of new arrival business tourists.