Thiago Cordeiro

This paper focuses on a decision support system that leverages artificial intelligence to evaluate the editorial quality of textbooks distributed to Brazilian public schools through the Brazilian Textbook Program. The program is crucial for enhancing educational resources for students and educators throughout Brazil. Our study relies on a dataset comprising images presenting a subset of textbook image characteristics to tackle a multiclass problem, categorizing them into three classes: sharp, defocused-blurred, and motion-blurred. We conducted experiments using the Fourier and Haar transforms to evaluate their impact on the performance of our CNN classification model. By using Fourier and 10-fold cross-validation, the mean accuracy of our models varied from 75.44% to 80.25%, the mean precision varied from 78.38% to 83.87%, the mean recall varied from 76.05% to 80.37%, and the mean Area Under the Curve (AUC) varied from 90.27% to 92.14%. By using Haar, the mean accuracy of our models varied from 69.08% to 87.14%, the mean precision varied from 78.38% to 87.80%, the mean recall varied from 76.05% to 87.25%, and the mean AUC varied from 90.27% to 93.95%. These results indicate that employing CNNs with data preprocessed using the Haar transform method can achieve more consistent results for assessing textbook image quality.

This study aims to investigate the psychometric parameters of the revised version of the Relational Factors for the Risk of School Dropout Scale (IAFREE-R). This updated version offers a methodological advantage over the original, as its design allows the scale to be completed solely by students, eliminating issues commonly associated with third-person measures (the case of teachers answering about the student's risk of dropping out of school). Several procedures were conducted, including item adaptations for self-report and psychometric procedures encompassing structural equation modeling and item response theory analysis. First, evidence of validity based on content was achieved across countries. Further, internal structure evidence indicated that IAFREE-R performed better than the original model tested in Brazil. In sum, the IAFREE-R ensured its operational advantages with psychometric evidence.

School dropout is a critical challenge for Brazilian education, having a multicausal nature and the potential to disrupt academic trajectories. This paper contributes to discussions on school dropout by covering primary and secondary educational levels. We conducted an extensive exploratory analysis of data provided by the National Institute of Educational Studies and Research Anísio Teixeira (Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira - INEP) from 1996 to 2022, considering factors such as government transitions and unforeseen events (e.g., financial crises and the COVID-19 pandemic). Thus, we handled millions of instances from the INEP database to explore data using Google Colab. We focused on the relationship between student dropout rates and various other factors, including categorical variables such as school location, type of administrative management, geographical region, and year, as well as numerical variables such as the number of classes, number of teachers, number of approved students, and number of failed students. Our goal was to identify which combination of these characteristics contributed most to the increase in student dropout rates. Our findings evidence that schools in urban areas of theNortheast region, whether municipal or state-run, have the highest student dropout rates in basic education in Brazil. The state of Alagoas shows a more pronounced duality in its education networks than the rest of Brazil, as the highest rates in elementary education occur in municipal schools. In contrast, state schools are the major contributors to high numbers of dropouts. Alagoas faces the most concerning dropout rates in high school education. The results showed periods of growth and decline in student dropout rates, aligning with government transitions, certain financial crises,and the COVID-19 pandemic. There is also a correlation between dropouts and failures in both elementary and high school education, with the correlation being much stronger in elementary education.

The increasing complexity of AI-based systems poses significant challenges in software engineering education, requiring an integrated approach that combines AI techniques with software development methodologies. Traditional educational models often fail to bridge this gap, limiting students' ability to apply theoretical knowledge to real-world AI-driven software solutions. This paper proposes the design of an Academic Software Factory (ASF) tailored for AI-based systems development within undergraduate software engineering education. The proposed ASF integrates Competency-Based Learning, Bloom’s Taxonomy, and Agile Scrum methodologies to provide a structured and industry-aligned learning experience. The ASF framework is developed through a four-phase methodology: (i) identifying characteristics and roles in AI-based system development, (ii) designing the ASF curriculum, (iii) defining a teaching methodology aligned with AI-specific software engineering challenges, and (iv) evaluating the ASF design through expert validation. A survey was conducted with ten professors specializing in software engineering and AI for preliminary validation. The feedback highlighted the ASF's potential to enhance educational outcomes, suggesting refinements such as assigning specialist roles to faculty members, reducing redundancy in pre-existing content, and emphasizing developing and presenting a minimum viable product as a key learning outcome.