Sentiment Analysis of Sirekap Application Review Using Logistic Regression Algorithm

Abstract

General Elections (Pemilu) is one of the crucial moments in democracy to elect representatives of the people. The General Elections Commission (KPU) launched the Sirekap application as an aid in the election process. This application allows polling station officers (KPPS) to record the vote count electronically. However, there have been some complaints and feedback from the public regarding the Sirekap application. To understand public sentiment towards the Sirekap application, this study was conducted by analyzing user reviews on the Google Play Store. The Logistic Regression algorithm is used to classify review sentiment into positive and negative. The analysis process involves data preprocessing, z-score normalization, dividing the data set into 80% training data and 20% test data, weighting words using the TF-IDF method, training the model using the Logistic Regression algorithm, and testing the model with a confusion matrix. The results of the analysis show that the Logistic Regression algorithm is effective in classifying the sentiment of the Sirekap application reviews with an accuracy of 91%. The precision score for the positive and negative classes are 90% and 92%, respectively. The recall score for the positive and negative classes are 94% and 87%, respectively. The f1-score for the positive and negative classes are 92% and 90%, respectively. The results of this sentiment analysis can also be used by the KPU to understand the level of user satisfaction and improve the quality of the Sirekap application for the 2024 Regional Head Elections (Pilkada).

Keywords

• analisis sentimen
• logistic regression
• sirekap