Automatic Analysis and Classification of Discourse Domains in Brazilian Portuguese
Abstract
This paper addresses the identification of the Juridical, Entertainment, Journalistic, Virtual, and Instructional discourse domains of Brazilian Portuguese at the sentence level, sampled from the Carolina corpus. We evaluate grammatical, lexical, and semantic properties. We demonstrate that the domains are discernible and organized into a consistent scale, which we associate with the oral-involved vs. literate-informational distinction based on comparison with other works. We trained Transformer classifiers on a new sentence dataset for domain identification, achieving high performance. The models' error patterns correlate with the identified scale, suggesting the models captured this dimension of variation. The datasets and models developed in this study are publicly available.
Copyright (c) 2025 Felipe Ribas Serras, Miguel de Mello Carpi, Mariana Lourenço Sturzeneker, Mayara Feliciano Palma, Aline Silva Costa, Vanessa Martins do Monte, Cristiane Namiuti, Maria Clara Ramos Morales Crespo, Maria Clara Paixão de Sousa, Marcelo Finger
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
