Klasifikasi kepuasan mahasiswa matematika UNP terhadap kualitas pelayanan Go-Food pada Gojek dengan metode naïve Bayes

Abstract

The development of Go-Food services in the surrounding community, including UNP Mathematics students, has caused various reactions, namely satisfaction and dissatisfaction with the services provided. Several factors are thought to result in Go-Food services being less than optimal based on the opinions of several experts, namely reviews of prices, payments, promotions, driver performance and suitability of specifications. Based on the results of distributing research questionnaires, there are several factors that cause students to be satisfied using Go-Food services and some are dissatisfied. The field of data mining science that will help companies to overcome this problem is classification techniques. Classification techniques in data mining will produce a classification model obtained from input in the form of training data, which has class variables. The classification model will map data object X to one of the previously defined classes Y. The classification method used is Naïve Bayes, which is defined as a combination of naïve and Bayes' theorem and produces the assumption that one independent variable is independent of each other. This research uses 44 training data and 44 test data. This classification presents the data into 50% training data and 50% testing data. The results of the classification of UNP Mathematics students' satisfaction with the quality of Go-Food services at Gojek using the naïve Bayes method obtained an accuracy of 86.3% and an APER value of 13.3%. This means that the naïve Bayes classification results are in the good classification range, which is concluded as good classification results.

Keywords: Classification, go-food, service quality, naïve bayes, APER.
MSC2020: 62C10