Classification On Category Of Public Responses On Television Program Using Naive Bayes Method

Abstract

Television has high popularity in Indonesia. This matter create a competition to get the highest rating. so that, sometimes the quality of content rarely noticed. This problem, make a lot of public responses likes from official website Komisi Penyiaran Indonesia (KPI), Short Message Service (SMS) or in social media. According that case, this paper using naive bayes method to categorize public responses from two social media likes Twitter and Facebook. This approach inspired from sentiment analysis which has been done before to television program based on public opinions from twitter using naive bayes method. There are several stages in this research start from taking data process from social media based on public opinions, pre-processing data, saving the data to MySQL database, and classification category process with naive bayes method. The researcher process the data responses become two main categories such as station tv category and program category. Output of this apps is list of information about public responses which has been categorized to make easier for KPI or the people to access also choose and determine the best program for their family. From 326 amount of data which has been used as dataset with precentage 80% of training data and 20% of testing data produce an accuracy value of 82%.