A new method for OTAs to analyze and predict users’ online behavior patterns and preferences

Abstract

OTAs use traditional statistics models to analyze huge number of data that is generated by the users, and then make predictions on their behavior patterns and preferences. In online tourism industry, statistics models were deeply studied and widely used by many big business companies. However, the author proposed a new model that is able to “freely” analyze and predict users’ online behavior patterns. It constructed a framework that is able to build and modify behavior models for different users in varied environments. Every single action conducted by the customer initiates an update of the existing model, which generates changes of parameters in quantity and/or quality. These changes get fed back to the environment where the customer faces, and cause changes of environment parameters in quantity and/or quality. The changes of the environment parameters stimulate another update cycle of the behavior model. In addition, the results of this study, which were based on the real data, demonstrated how users’ booking behavior affects the OTA’s products recommendations.