IMPLEMENTATION OF TEXT MINING FOR CLASSIFYING COMMUNITY QUESTIONS VIA WHATSAPP WITH THE NAÏVE BAYES CLASSIFIER METHOD

Abstract

Text mining is the process of exploring knowledge based on specific patterns of textual data retrieval. There was an increase in the amount of text data from community questions on the Whatsapp information service at Surabaya Immigration Office which can be processed into detailed and complete information. Text data entered through Whatsapp question has also not been classified specifically, structured and also has not been published. This study aims to explain the characteristics of incoming messages provided by the public through the Whatsapp information service and to explain the process of classifying community questions according to the field of immigration public services namely WNI and WNA. The authors used the classification method with the Naïve Bayes Classifier (NBC). Obtained the value of classification accuracy with algorithms and methods using the Naïve Bayes Classifier on the training data equal to 93.5% and testing data equal to 95% that included in the excellent scale. Therefore, Naïve Bayes Classifier method is very well applied for classifying public questions and SIPESAN system.