Esquema de clasificación de información universitaria basado en NERC: caso noticias UTPL

Abstract

This research proposes the development of a classification scheme of college information. Natural Language Processing (NLP) and Machine Learning (AA): two methodologies for development based on the recognition and classification techniques Named Entity (NERC), called apply. NERC allow harness the wealth of the context in which the named entities such as are presented: people, organizations, locations, dates, and titles of person. As first phase of the creation of a corpus of two hundred news, labeling manually, the same as used for the analysis and pattern making. In second phase is PLN methodology, where the GATE (General Architecture for Text Engineering) tool is used, is an open-source Java-based infrastructure serves to develop and reuse software components to solve the problem of classification, and The last phase is the methodology of machine learning, where the SVM algorithm and Support Vector Machines classification, for which the results are presented through a demo applies.