Knowledge Discovery from Trouble Ticketing Reports in a Large Telecommunication Company

This paper describes the work developed byTelefónica I+D about an application of advanced DataMining, Text Mining and Machine Learning techniquesfor the study of the network elements failures managedby the Trouble Ticketing System of a largetelecommunication company, in order to be able toanalyze, prioritize and, in some cases, solve withouthuman intervention the huge amount of trouble reportsto be managed. Furthermore, this paper will presentthe techniques used for its achievement, as well as theresults obtained so far, showing how these techniquesmay help important companies to save plenty of timeand resources in fault management, improving theservice quality.

1. Introduction

Traditionally, important companies have jealously stored their business data, using them only for reporting and auditing. However, in recent years companies have started to realize that the huge amount of available information offers a much broader scope. That is the reason why Data Mining techniques are quickly expanding, providing companies with key information. The present research wants to take a further step, with the aim not only to get relevant information, but also to use this information in order to improve and optimize the Business Processes of a Telecommunication Company.