Metodo de procesamiento de lenguaje natural y tecnicas de mineria de datos aplicadas a la clasificacion de incidentes informaticos

Abstract

This article presents a methodology that applies natural language processing and classification algorithms by us­ing data mining techniques, and incorporating procedures for validation and verification of significance. This is conducted according to the analysis and selection of data and results based on quality statistical analysis, which guarantees the effectiveness percentage in knowledge construction. The analysis of computer incidents within an educational institution and a standardized database of historical computer incidents collected by the Service Desk area is used as case study. Such area is linked to all information technology processes and focuses on the support requirements for the performance of employee activities. As long as users’ requirements are not fulfilled in a timely manner, the impact of incidents may give rise to work problems at different levels, making it difficult to plan or prevent incidents resolution due to their unforeseen nature.