Machine Learning-based Automated Problem Categorization in a Helpdesk Ticketing Application
Albert Miliano, Ignatius Steven, Krisananda Putera Kosim, Riyanto Jayadi, Tuga Mauritsius · 2020
To handle internal complaints, some companies have implemented a helpdesk system while some other processes are still carried out manually. The helpdesk ticket is categorized manually by human operators. This procedure prone to be an error resulting in many tickets bounced to the wrong business unit and delaying the complaint handling. In this study, an automated problem categorization based on the title of helpdesk ticketing using machine learning is studied. The results show that the random forest classification has the highest accuracy value of 82%.