Brazil-Portugal comparative study of journalistic stances on covid-19 vaccination using natural language processing

ABSTRACT

This paper reports an investigation versed in identifying journalistic stances through the utilization of algorithmic models based on natural language processing. To this end, a comparative case study between Brazil and Portugal is presented, focusing on the identification and analysis of digital news portals regarding political issues associated with covid-19 vaccination campaigns in the years 2021 and 2022. Five journalistic outlets from each country are considered based on popularity criteria. The methodological approach is grounded in the use of computational tools for data extraction and analysis, particularly within the scope of natural language processing (NLP), in order to label the texts according to expressed stances (sentiment analysis). The results reveal an increase in negative sentiment in Brazilian newspapers and a scenario of balance in the debate outlined in Portuguese newspapers.

KEYWORDS
journalism; sentiment analysis; vaccination; covid-19