Um Estudo de Caso do Uso de Mineração de Dados e Aprendizado de Máquina no Aprimoramento de Inspeções de Estações Rádio Base

Abstract

Currently, phone companies spend workforce in a sluggish process of radio base stations inspections. Having as background the emergence of the Industry 4.0, to incorporate computational intelligence algorithms in the speeding up of this process figures as a competitive advantage. It is in this context that this work presents an algorithmic solution with the objective of helping telecommunications technicians and engineers in the task of determining which inspection items are dispensable. The necessary information to train machine learning algorithms to suggest to users which items have the highest probability of dispense was extracted through the application of data mining and natural language processing tools. This work, therefore, represents a preliminary effort in the acceleration of those inspections.