Predicting Tourists’ Behavior Of Virtual Museum Using Support Vector Machine With Feature Selection Technique

Abstract

Multimedia technology has greatly influenced tourism industry in Thailand. It can increasingly promote Thai cultural tourism byapplying the virtual reality technology to create Chaiya national virtual museum. Significantly, the virtual reality method is considered as a modern and popular method integrating with a game design. The paper employs the method that selects a variety of information features and the appropriate characteristics from support vector machine classification technique in order to create a model to predict tourists’ behavior. The researchers adopt the feature selection methods consisting of the genetic search, the best-first search, and the greedy stepwise search algorithms integrated with the wrapper method. These methods are applied to evaluate the performance of Chaiya national virtual museum. In sum, the results of this work reveal the feature selection textbf{{method can select the best data set features and improve the efficiency of predictive model.