Mobile-based recommendation system for the tour package using the hybrid method

Although there has been a lot of research on traveler recommendations, there have been no studies that discuss the recommendations of tour packages from travel collections in the past. Therefore, in this study it is important to do a related study 1) Development of a mobile-based recommendation system using the Hybrid Method. 2) Accuracy testing of the system in providing tour package recommendations. This study uses the CBR method to provide tour package recommendations based on past travel data. There are 4 stages of the process, such as: Retrieve, Reuse, Revision, and Retain. In this study, focused on the Retrieve stage using a hybrid method. The hybrid method on the mobile-based recommendation system is a combination of the methods of Naive Bayes, Bayes Theorem, and Dempster Shafer. Where Naive Bayes is used to calculating probabilities such as age and frequency of visits. The Bayes Theorem is used to calculate probabilities such as country, gender, and destination. To determine the mass value of a combination of evidence, then use the Dempster Shafer method. Based on the accuracy test, it was affirmed that the total accuracy of the tour package by the mobile-based recommendation is 95%. While the error rate of the system in providing tour packages is 5%. © 2018, International Association of Online Engineering.