Rules mining based on clustering of inbound tourists in Thailand

Abstract

Tourism industries are growing up rapidly with more competition. So, travel agencies or tourism organizations must have a good planning and provide campaign for tourist’s needs. This study proposes the usage of data mining for tourism industries in Thailand. Data clustering and association rule mining were chosen as the data mining methods in order to discover useful knowledge. Two-level clustering with decision tree bagging was applied to construct the segments of tourist. Apriori algorithm was then used to find the rules on each cluster. The experimental results indicated that the tourists data was separated into eleven differently segments and decision tree bagging for attributes weighting can enhance the quality of clusters. The eleven segments were analyzed in order to identify tourists’ behavior patterns and their preferences. Association rule mining was applied to each segment in order to find the relationship among the features of tourist data. The rules were filtered again by experts. The clustering and association rule results can be served to tourism organization in order to support their strategic and market planning.