Selection System of Madura Tourism Object as The Best Recommendation Based on Comparative Analysis of K-NN and Naive Bayes Methods

Abstract

The current decision-making system is developing very rapidly. Various methods can be used to assist recommendations in the decision-making process. The technique chosen in decision-making is sometimes irrelevant and less precise. In addition, differences in classification methods often result in different results in the recommendation process. Therefore, to determine the decision-making, the decision support system is one of the tools in solving this problem. This study proposes an approach to help make decisions as the best recommendations in choosing tourist attractions in Madura by comparing the classification methods, namely the K-Nearest Neighbor (K-NN) and Naive Bayes (NB) methods. For data validation, k-fold cross-validation is used. The NB method was chosen because it can produce maximum accuracy with little training data. In comparison, the K-NN method was selected because it is robust against random data, and this method is the simplest. The performance of the two ways will be compared to determine which method is better in the selection process. This research aims to help tourists at home and abroad choose suitable tourist attractions according to the tourists’ needs. This study uses 21 tourist destinations. The parameters used for the tourist attraction selection process consist of 11 indicators, including gender, age, occupation, education, marital status, tourism type, management services, facilities, education, ticket prices, and sales trends. From 320 datasets with several trials, it produces the highest accuracy value in the K-NN method with the 1st fold (K1), which is 77.81%, while the NB method has the highest accuracy value in the 2nd fold (K2) is 53.75%.