KOMPARASI ALGORITMA KLASIFIKASI TEXT MINING PADA REVIEW RESTORAN

Abstract

At this time, where the development of technology is developing very rapidly, and everyone has the right to express his opinion on a matter. One of them is conducting a review of a restaurant. The review, can be created from food, decoration, or service. This, is used by business people to find out consumer ratings about the restaurants they manage. However, the review data must be processed using the right algorithm. Then this research is conducted to find out which algorithm is more feasible to use to get the highest accuracy. The method used is Naïve Bayes (NB), and k-Nearest Neighbor (k-NN). From the process that has been done, it is obtained that the accuracy of Naïve Bayes is 75.50% with a Kappa value of 0.510, and the accuracy results when using the k-Nearest Neighbor algorithm is 89.50% with the AUC value of 0.790. The use of the k-Nearest Neighbor algorithm can help in making more appropriate decisions for hotel reviews at this time, because the resulting accuracy is greater than the Naïve Bayes Algorithm.