A Comparative Study for Predicting the Product Type Based on the Characteristics of the Customer Complaint for Orange Jordan Telecommunication Company

This paper aims to investigate Telecom's products that customers complain about. Telecom companies can use customer complaints to improve their services and loyalty through insightful analysis of complaint categories. The dataset used in this paper was obtained from a well-known company in Jordan named Orange Jordan. The dataset contains complaints received from several clients at the company. The dataset has undergone several preprocessing steps, including cleaning, encoding, and feature selection. Several machine learning classifiers were used to classify the complaints by product type, such as Internet or mobile products. Based on the classification metrics, the Decision Tree model was shown to be the reliable model that can accurately predict the product type for the selected dataset, where it performed slightly better than the KNN model and outperformed the GussianNB model. To enhance the performance of the classification model, it is recommended that the data retrieved from the company system be improved to avoid missing values occurring in the dataset.