Classification of Comments Regarding the Politeknik Caltex Riau Using the Naive Bayes Algorithm from Instagram Social Media

Abstract

Instagram Politeknik Caltex Riau is one of the social media used by the Riau Caltex Polytechnic for media promotion of various activities, achievements and other matters related to the Caltex Riau Polytechnic. Instagram Politeknik Caltex Riau was created in 2015. Of the various posts on the @politeknikcaltexriau account, there are many comments that are critical, input, and questions. So that these comments can be used by the Riau Caltex Polytechnic by classifying comments based on the categories needed by the Caltex Riau Polytechnic. Therefore, we need a system that can perform the comment classification process using text mining that utilizes the Naive Bayes algorithm. With this system, it is expected to know the percentage and pattern of public comments on the @politeknikcaltexriau account per category and the trend of comments in certain months. Based on the results of tests carried out on the classification system that was built, using blackbox testing, it was found that 100% of the system's functionality was running well. Furthermore, text mining techniques and the Naive Bayes algorithm have been successfully adapted to this system in categorizing Instagram comments on the @politeknikcaltexriau account. The system has been tested for functionality and has an average accuracy of 72.911%. Keywords: comment classification, text mining and naïve bayes