Identificar sentimientos en cuarentena por la covid-19 mediante clasificador léxico y aprendizaje supervisado

The Covid-19 began to affect Peru on March 6 of 2020, preventive measures were started to prevent the spread. On March 15 compulsory social isolation began throughout Peru, the people use Twitter to exchange various information about social isolation, this is important for authorities and the public because it helps to consider strategies to avoid contagion. The present work has the objective to classify the positive and negative sentiment that were expressed on Twitter through the proposal of the Lexical Word Classifier and the use of classifying algorithms. The result obtained was that the most frequent words are: Quarantine, Covid and Home. The positive words were Good and Win, the negative word was Strange. The sentiment classification model reached 91.5% accuracy using the Support Vector Machine algorithm and the Lexicon Word Classifier.