Ferramenta para auxílio à análise de evasão através de intervalos de predição de séries temporais

Abstract

University dropout represents a recurring challenge in the context of higher education institutions, requiring technological solutions that support the monitoring and organization of institutional information. In this context, this work aims to develop the web-based tool called Oráculo, designed to support the exploration, visualization, and organization of data related to academic dropout, with the goal of facilitating access to information and supporting decision-making processes. The solution was conceived based on a modern, scalable, and decoupled architecture, grounded in the client–server paradigm. In this approach, the backend acts as a service provider through a RESTful API, while the frontend is responsible for user interaction, enabling independent evolution of system layers and easier system maintenance. The backend was developed using the Flask microframework within the Python ecosystem, while the frontend was implemented with Next.js and React, ensuring a responsive and reactive interface suitable for modern web browsers. Data persistence is handled through a lightweight relational database, with access abstraction provided by an object–relational mapping layer. The tool was designed to operate in cloud computing environments and adopts modern software engineering practices, such as version control, continuous integration, and continuous delivery. As a result, Oráculo stands out as a modular and extensible solution aligned with the objectives of this work, contributing to institutional support for monitoring university dropout through an integrated technological approach.