Classification of Complaint Categories in E-Commerce: A Case Study of PT Bukalapak

Bukalapak is one of the companies in Indonesia that is engaged in E-Commerce. It was recorded that in 2019, Bukalapak experienced a 230% increase in user growth compared to the previous year. Unfortunately, the growth in users is also followed by an increase in the number of complaints that occur in Bukalapak. It was recorded that in 2020, Bukalapak complaints increased by 50% at the end of 2020 when compared to the beginning of 2020. The increase in complaints led to an increase in the average handle time for complaints which led to a decrease in user satisfaction. 3 main issues that cause an increase in average handle time, namely an unstable system, complaint categorization is slow, and difficult to find solutions. This is certainly a concern for the management. In this study, the classification of complaints categories in Bukalapak will be carried out. This study aims to find out what classification model is suitable to be used in determining the category of complaints in Bukalapak. The classification model that will be used in this research is Logistic Regression, k Nearest Neighbor, and Support Vector Machine. While the data that will be used is data on complaints from January 2021 to December 2021. From the results of the study, it was found that the logistic regression classification model had the highest value among the other 2 models. The logistic regression model managed to get an accuracy value of 83.9%. The second position is occupied by the k Nearest Neighbors model with an accuracy of 77.6%. Last occupied by the SVM model with an accuracy value of 40.5%.