Sentiment Analysis of Office Automation Application in One of Indonesian Ministries

The Office Automation application is a common application in one of the ministries in Indonesia that contains features of correspondence management, task management, collaboration tools, attendance, and meeting management. The application is expected to be used for all Echelon I Units and helps smooth business processes at the Ministry during the pandemic. Unfortunately, not all Echelon I Units use the application because the application still requires many improvements and additional features according to the needs of the Echelon I Unit. Based on this, sentiment analysis of the Office Automation application was carried out based on user reviews using text mining techniques with Naive Bayes classification, SVM, K-NN, Logistic Regression, Decision Tree, and Random Forest. The analysis is done by looking at the evaluation results in accuracy, precision, and recall and based on the Feature and Post Tagger combination, which produces the best evaluation value in the Logistic Regression model. In addition, the results of the built model also analyze reviews from Office Automation users in it. In recommendations, positive and negative reviews are assessed from 5 aspects: User Interface, User Experience, Functionality and Performance, Security, Support, and Update.