Implementasi Naïve Bayes Classifier untuk Prediksi Jenis Pelayanan Berdasarkan Preferensi Klien Berbasis Web Pada PT. Karya Rans Indonesia

Abstract

PT Karya Rans Indonesia is a company engaged in brand activation and promotional services. The determination of promotional service types, which still relies on staff experience, has resulted in the underutilization of clients' historical preference data. This study aims to implement the Naïve Bayes Classifier method in a web-based prediction system to predict promotional service types based on client characteristics. The Naïve Bayes Classifier method was implemented using the Laplace Smoothing technique to calculate classification probabilities. The results show that the developed system is capable of predicting promotional service types based on promotion objectives, activity types, target audiences, and sponsorship attributes. The classification results indicate that the Brand Awareness class achieved the highest posterior probability of 0.0112 (47.29%), and was therefore selected as the prediction result. The proposed system is expected to support a more objective, consistent, and data-driven process for determining promotional service types.