ABSTRACT
Automatic syntactic parsers can support the teaching of Brazilian Portuguese. However, existing corpora that underpin such parsers are primarily designed for industrial applications rather than for the instruction of the vernacular. In this context, we propose the Araucárias corpus, a collection of syntactic trees derived from well-formed Portuguese sentences. The aim is to demonstrate its relevance for educational purposes in comparison to other existing corpora. Due to its relatively intuitive, simple, and evocative nature—attributes that are absent in the evaluated existing corpora—Araucárias is expected to offer a more viable approach to teaching Portuguese syntax.