• Ahmad Syahrul Mubarok Program Studi S1 Teknik Informatika,Fakultas Teknologi Informasi Universitas Hasyim Asy’ari Jombang
  • Dedy Rahman Prehanto Fakultas Teknologi Informasi, Universitas Hasyim Asy’ari Tebuireng Jombang
  • Mahrus Ali Fakultas Teknologi Informasi, Universitas Hasyim Asy’ari Tebuireng Jombang


The need for communication technology that can find out the customer's buying interest is something that is
needed by the company, the goal is that the company can obtain consideration regarding product sales and
appropriately take a policy regarding pricing of a product, the amount of product supply and so on. The naïve
bayes classifier method is sufficient to be able to find out how much the customer's buying interest in an
internet package product is by analyzing the results of past sales, meaning that the Naive Bayes algorithm
can predict future opportunities based on past experience, by defining each class from all attribute. The naive
bayes classification is assumed that there is or not a certain feature and a class has nothing to do with other
characteristics and classes. This study defines several characteristics, namely operator, quota, active period
and product price. From the results of this study is the author can find the value of accuracy from the results
of calculating interest and not by comparing the results of predictions with actual results outside the training
data. Based on data on internet package product sales as training data, the Naive Bayes method successfully
classifies 10 product data with 858 transaction data as training data and 10 product data with 115 transaction
data as test data. From the analysis of all test data and training data, the truth ratio is 8/10, which is 8 out of
10 products with predictive values that are of true value. So that the accuracy value is 80% and to find out
the customer's buying interest by calculating the sales data that has been past with the provisions and in
accordance with the specified label, the results of this prediction are expected to be a better policy
consideration for the sales business.
Keywords: naïve bayes classifier, sales of internet package products