Machine Learning Master's Thesis
01.10.20192 Min Read — In Learning
Authors:  Miruna Lazar, Jose Diaz

After spending a summer as an intern in Telia Norway, I very quickly understood the impact of the company on the telecommunication industry in Scandinavia and the great things the company is set to achieve. For this reason, I hoped very much to get a chance to contribute to the great developments that the company is currently working on, by conducting together with my thesis partner, our thesis project on a case given by Telia.

So who are we? We are two Master of Science in Business Analytics students, from BI Norwegian Business School. Starting with this fall and lasting until summer 2020, we will be working within the IT Analytics department to develop a machine learning model that will accurately categorize the inquiries that GET customers have, via various sources such as chat, telephone or email. We will be closely supervised by our Predictive Analytics and Machine Learning professor throughout this project.

Today, the classification of the inquiries is sometimes done manually, by the customer service consultant, when he or she receives it. Very often, this manual classification is inaccurate or is missing, and it leads to a wrong root cause analysis. By creating a classification model, the company will be able to improve the process of keeping track of all the different issues that customers are facing and enhance the root cause analysis, with the aim of significantly decreasing the number of inquiries per year. In the same time, this classification model can set the foundations for a future development of a chatbot, that can further automize the process.

For developing this model, we will be applying Natural Language Processing (NLP) algorithms for reading, analysing and understanding the body of the texts or comments. NLP is an emerging technology, which is at the foundation of many forms of AI that we are used to seeing nowadays. Technologies that use NLP are increasing at a very high pace and are facing continuous development. This represents a topic that we both are very passionate about and are so excited and looking forward to be able to apply it in a real world case, and see how it will positively impact the user experience for GET customers.

Thank you,

Miruna and José

PS. If you want to learn more about Natural Language Processing, here is a nice place to start. DS.

Miruna

Jose