Análisis de la información de foros en cursos MOOC mediante técnicas de procesamiento de lenguaje natural

Abstract

Abstract: Online learning is one of the characteristics of this millennium, due to the large amount of information that is generated daily and the accessibility of devices to access the web. Based on this change in the learning perspective, several institutions worldwide have conceived the new paradigm of online teaching as an opportunity to globalize their programs. Thus, for a few years now, courses, seminars and even university courses have been taught through the web, which has generated a large amount of interaction between teachers and students located in different parts of the world, an interaction that creates a massive amount of data that have not been explored, so this project intends to carry out such exploration using artificial intelligence techniques. In the present work, the extraction of messages in forums that are generated from student interaction in the massive open online courses (MOOC) of the Udacity platform (using scraping techniques) is carried out, in order to rescue text patterns. by using natural language processing techniques (such as entity recognition and n-gram analysis) and machine learning techniques (using the Naïve Bayes algorithm)