Fonte: Revista de Turismo e Identidad; Vol. 2 Núm. 1 (2021); 55-71
Palavras-chave:
Resumo: Activation of tourism is one of the key subjects for the airline industry. Internet contains a lot of information about tourists. This paper aims at analyzing the opinion of the tourists who traveled by certain South America airlines, using the sentiment analysis technique, employed in the study of their messages. The resource used for analysis is the information in twitter, provided by these airlines customers. First, a method for extracting published phrases related to target locations and 'hashtags' was presented. Then, it was analyzed the polarity of the tweets extracted; creating positive, negative and eventually neutral opinions. In this process, there was utilized an unsupervised learning technique using seed words. The experimental result on the classification shows the efficacy of the applied method. Preliminary (descriptive) results as well as the basic proposal for a predictive model are herein attached.
