Monitoring of Information and Sentiment Analysis of Brands on Social Media in the Luso-Brazilian Context

Abstract

The advancement of digital technologies and the popularization of social media have profoundly transformed how brands interact with the public, making strategic information management in this environment essential. This study aimed to analyze the scientific production related to sentiment analysis of brands on social media, with an emphasis on the Luso-Brazilian context. The goal was to provide an overview of the volume of publications, main application sectors, pursued objectives, achieved benefits, and challenges faced by organizations.

The research adopted an exploratory-descriptive and qualitative-quantitative approach. Data collection was carried out through a Systematic Literature Review (SLR) in the Scopus database, following the protocol of Dresch, Lacerda, and Antunes Jr. (2015). The inclusion criteria focused on publications from 2008 to 2024, in Portuguese and English, related to sentiment analysis in social media within the Luso-Brazilian context.

The search yielded 14,929 documents, from which, after applying exclusion criteria, 356 studies were analyzed. The results revealed significant growth in scientific output on the topic, especially between 2018 and 2022, although Brazil and Portugal showed fluctuations in publication rates. Globally, India, the United States, and China lead in sentiment analysis publications, while Brazil and Portugal hold intermediate positions, with 254 and 129 publications, respectively.

The analyzed publications are mostly concentrated in the fields of Computer Science and Engineering, but also include Business, Social Sciences, and Medicine, highlighting the growing interdisciplinarity of sentiment analysis. The main objectives found in the studies were: evaluating brand perception and reputation (57 studies), supporting marketing strategies (25), detecting health risks (23), and predicting consumer or electoral behaviors (16).

The benefits obtained through sentiment analysis were diverse: better understanding of public behavior (44 studies), support for image and reputation management (30), more effective crisis response (22), and anticipation of market trends (17). These results highlight the usefulness of the technique as a strategic tool for organizations seeking to enhance their competitiveness and responsiveness in the digital environment.

On the other hand, the main challenges faced by companies include difficulties in interpreting irony and sarcasm (23 studies), low-quality or noisy data (25), and sampling representativeness issues (20). These technical and methodological barriers emphasize the need to combine automated tools with human analysis to ensure greater accuracy.

The final considerations point out that sentiment analysis is not an isolated technique, but rather a concrete application of information management principles in the digital context. By transforming unstructured data into strategic insights, it directly supports organizational decision-making. Despite the growing scientific production in Brazil and Portugal, the study acknowledges limitations such as the restricted geographic scope, use of a single database (Scopus), and lack of detail on the specific technologies employed in the analyzed studies.

For future research, it is recommended to broaden the scope to include other countries and databases, as well as to deepen the investigation into the tools and algorithms used. It is also suggested to explore practical applications of sentiment analysis in underexplored sectors, expanding the knowledge frontiers in this emerging field.