Ecuadorian Higher Education in COVID-19: A Sentiment Analysis


Abstract:

COVID-19 is a major medical problem worldwide, but it also leads to educational problems. The aim of this work is to contribute with information about the feelings generated in university students and to know the main characteristics of Higher Education in Ecuador in times of pandemic. Specifically, the question What are the feelings of university students in times of the COVID-19? was answered. For this purpose, a quantitative, transversal and non-experimental research was carried out. Fifty-five unstructured anonymous interviews were conducted. It was applied to students from 16 Higher Education Institutions. Feelings were analyzed using techniques such as Latent Dirichlet Allocation (LDA), and through the computer programs MATLAB and NVIVO applied to the 500 phrases, obtained from 55 interviews with an average of 47 words per phrase. When carrying out the sentiment analysis, it was obtained that approximately 64% was negative, 11% neutral and 25% positive. LDA found that the 55 interviews were explained by 2 unobserved groups represented by the word clouds of topics 7 and 14. The unobserved groups show feelings such as stress, tiredness, problems and effort that may be related to the words people, evaluation, type, education and strength. This research can be complemented with a study that allows to deepen the type of negative, neutral and positive feelings and to determine their possible causes.

Año de publicación:

2020

Keywords:

  • ECUADOR
  • covid-19
  • higher education
  • matlab
  • sentiment analysis
  • NVIVO

Fuente:

scopusscopus

Tipo de documento:

Conference Object

Estado:

Acceso restringido

Áreas de conocimiento:

  • Educación superior

Áreas temáticas de Dewey:

  • Educación superior
  • Educación
  • Otros problemas y servicios sociales
Procesado con IAProcesado con IA

Objetivos de Desarrollo Sostenible:

  • ODS 4: Educación de calidad
  • ODS 10: Reducción de las desigualdades
  • ODS 3: Salud y bienestar
Procesado con IAProcesado con IA