UM TEXT CLASSIFICATION PROCESS

SENTIMENT ANALYSIS OF OPINIONS ON TRIPADVISOR ABOUT THE ATTRACTION OKTOBERFEST BLUMENAU

Authors

  • Marcio Crescencio Universidade Federal de Santa Catarina
  • Alexandre Leopoldo Gonçalves
  • José Leomar Todesco

DOI:

https://doi.org/10.48090/ciki.v1i1.867

Keywords:

Sentiment analysis; Machine Learning; Classification; Knowledge Discovery in Data process

Abstract

The aim of this article was to develop an experiment with a sentiment analysis of user opinions on TripAdvisor® about the Oktoberfest Blumenau attraction using Data Mining and Machine Learning through a Knowledge Discovery in Data process. Two supervised sentiment classification approaches were implemented in Python®, based-model probabilistic on the Multinomial Naïve Bayes and the vector representation of words model using Word2Vec. The performance of models was evaluated and compared with measurements: Accuracy, Precision, Recall and F-score. The probabilistic model achieved an accuracy of 90%, while the recurrent neural network model LSTM was 92%. The feeling of the opinions is positive for the features of traditional German party, variety and number of drinks, and music and animation. Opposite for the queues in bathrooms and overcrowding on Saturday nights.

Downloads

Published

2020-11-18

How to Cite

Crescencio, M., Gonçalves, A. L. ., & Todesco, J. L. (2020). UM TEXT CLASSIFICATION PROCESS: SENTIMENT ANALYSIS OF OPINIONS ON TRIPADVISOR ABOUT THE ATTRACTION OKTOBERFEST BLUMENAU. International Congress of Knowledge and Innovation - Ciki, 1(1). https://doi.org/10.48090/ciki.v1i1.867

Issue

Section

Full Paper

Most read articles by the same author(s)

1 2 > >>