Modelo para la clasificación de correos electrónicos de incidentes de la mesa de ayuda del Ministerio de Salud Pública planta central mediante minería de texto

Abstract

Abstract: Due to the high demand for emails received by the Ministry's Help Desk of Public Health of Ecuador, between requirements and incidents, it is necessary to have model that allows automatic classification, based on text mining and learning automatic in order to reduce response times and optimize the human resources that owns the institution. In this degree work, the use of text mining and analysis techniques is proposed. machine learning to classify emails based on their content which are sent by officials and citizens. Two methods are evaluated for the Text cleaning used in text mining and two machine learning techniques using the data received in the help desk. The results show that the algorithm of a neural network presents better performance, achieving 80% accuracy.