Expert recommendation for trouble ticket routing

Abstract

A trouble ticket is an important information carrier in system maintenance, which records problem symptoms, the resolving process, and resolutions. A critical challenge for the ticket management system is how to quickly assign a proper expert to deal with trouble tickets and fix problems. Thousands of tickets bouncing among multiple experts before being fixed will consume limited system maintenance resources and may also violate the service level agreement (SLA). Thus, for an incoming ticket, an expert should be recommended as quickly as possible in order to reduce the processing delay.In this paper, to address the challenge in the expert assignment, we exploit an expert collaboration network model by combining expertise profiles and social profiles learned from problem descriptions and resolution sequences of the historical resolved tickets, and develop several two-stage expert recommendation algorithms to determine a resolver to solve the problem. To evaluate the effectiveness of expert recommendation algorithms, we conduct extensive experiments on a real ticket data set. The experimental results show that the proposed algorithms can effectively shorten the mean number of steps to resolve (MSTR) with a high recommendation precision, especially for the long routing sequences generated from manual assignments. The proposed model and algorithms have the potential of being used in a ticket routing recommendation engine to greatly reduce human intervention in the routing process.