Desenvolvimento de uma hipermídia adaptativa para o estudo da língua latina

Abstract

The teaching of the Latin language is historically marked by a dichotomy between the GrammarTranslation method and the Reading/Direct methods, with both traditional approaches following a linear, "one-size-fits-all" model. This work’s central objective is to propose the design and modeling of an Adaptive Hypermedia System capable of reconciling these two pedagogical approaches by personalizing the learning path. To this end, the Lingua Latina Adaptativa solution was designed using the UWE software engineering methodology, which extends UML for hypermedia application design. The system was detailed in its three modeling phases: Conceptual Design, which defines the Domain Model (with grammatical concepts, texts, and exams), the User Model (tracking student progress), and the Adaptation Model (based on prerequisite rules); Navigational Design (defining contexts and links); and Presentational Design (modeling the interface). To materialize the modeling specification, a functional prototype was developed using the Next.js framework, integrated with a PostgreSQL database, and hosted on the Vercel platform. The prototype implements the adaptive dependency graph logic on a React Flow visual editor, employing Tarjan’s algorithm to ensure acyclicity and prevent pedagogical deadlocks. The system’s validation was carried out through a Usability Heuristic Evaluation based on Nielsen’s ten principles and an empirical user experience trial with a language teaching professional. The results attest to the platform’s ergonomic quality and the suitability of the adaptive learning path for the student. [resumo fornecido pelo autor]