Implementation of machine learning algorithms for the classification of online customer support tickets

Abstract

This master thesis presents a study on the use of machine learning techniques for the classification of customer support tickets. The objective is to develop a model that can accurately classify support tickets into predefined categories, based on the summary of the messages. The study begins with a thorough literature review on the existing methods and techniques for support ticket classification, including text preprocessing, feature extraction, and machine learning algorithms. Based on the literature review, a methodology is proposed, which includes data collection, preprocessing, feature extraction, model training, and evaluation. The proposed methodology is applied to a dataset of support tickets collected from a real-world customer support system. The dataset is preprocessed and labeled, and several feature extraction techniques are applied, including word embeddings and bag-ofwords representations. A variety of machine learning algorithms are evaluated, including logistic regression, decision trees, and convolutional neural networks. Experimental results show that the best performing model is a Logistic Regression model, which achieves a F1 score of 0.69. In contrast, GaussianNB and the second Neural Network implementation were less effective, highlighting the importance of selecting and fine-tuning models based on the specific needs of the task at hand. The proposed methodology and the experimental results demonstrate the feasibility and effectiveness of using machine learning techniques for support ticket classification. The developed model can assist customer support teams in efficiently managing and resolving customer issues, leading to improved customer satisfaction and reduced operational costs. Overall, this master thesis contributes to the field of natural language processing and machine learning, and provides insights and recommendations for future research on support ticket classification.