This complete article of the research category presents proposes the application of Educational Data Mining (EDM) techniques using data generated by a web-based Intel-ligent Tutoring System (ITS) developed for teaching algebra. The aim is to classify students' academic performance, seeking evidence to assist teachers in the decision-making process. To achieve this goal, we utilized the Knowledge Discovery in Databases (KDD) methodology to analyze students' academic data and identify patterns, trends, and relationships that can be used to enhance the teaching and learning process and make informed decisions in the educational context. Practical results revealed that the Random Forest algorithm classified students' academic performance in each evaluation period with an accuracy of approximately 94 %. These findings derived from data analysis can serve as a basis for decision-making, aiding the teaching process for teachers, tutors, and administrators, as well as monitoring academic performance and identifying challenges present in each class. In summary, these results provide a comprehensive guide to the use of EDM techniques in a specific context.

This research paper discusses the impact of teaching and learning technologies on generating extensive data in education. Despite this data, a significant portion remains unanalyzed, creating a gap in utilizing its potential insights. Educational dashboards address this issue by presenting student information and interactions in online learning environments, utilizing Educational Data Mining (EDM) techniques. EDM aims to understand student behavior, enhance teaching processes, identify learning patterns, and personalize content. It aids in performance analysis, helping struggling students and guiding teaching adjustments. The study proposes an educational dashboard for analyzing Intelligent Tutoring System (ITS) data, specifically in algebra teaching. Using decision tree algorithm to extract insights from student interactions. The developed web dashboard organizes and visualizes student performance data, aiding monitoring, decision-making, communication, and transparency. Educational dashboards enhance decision-making, data analysis, and communication, benefiting both students and the educational system's performance.

This full paper of the research category describes the ERPLab as an innovative educational platform that aims to promote inclusivity and diversity in programming and robotics. It provides a unique learning experience, particularly in remote or resource-limited areas, making scientific education more accessible to a broader audience. ERPLab offers students an understanding and integration of scientific knowledge by combining biology, engineering, and programming. ERPLab plays a significant role in educational technology research by exploring new possibilities for remote teaching and providing valuable data and insights to evaluate the effectiveness of this approach. One of the standout features of ERPLab is its use of an automated greenhouse controlled by a 3D printer-produced robotic arm equipped with sensors, and two Wi-Fi cameras, esp32cam, facilitating observation from multiple angles, allowing students to explore new avenues of scientific learning. This system enables students to learn about plant cultivation while developing their programming and technological skills. ERPLab combines advanced technology with natural repellent plants to promote innovative teaching and research methods. The plant selection within the ERPLab includes natural repellents like rosemary (Rosmarinus officinalis), rue (Ruta graveolens), and basil (Ocimum basilicum). These essential programming skills are directly applicable to the rapidly advancing INDUSTRY 4.0. Some preliminary work was carried out in a school environment, and the results of feedback from some students and teachers were promising regarding the proposed approach. For future work, we will carry out qualitative and quantitative research to observe the interest and development of users for the program of the platform in the context of regular classes.

This complete article of the research category presentes proposes the application of Educational Data Mining (EDM) techniques, based on data generated by a web-based Intelligent Tutor System (ITS), developed for the teaching of algebra, which helps students to develop fundamental knowledge of mathematics, in addition to carrying out a comparative study between the level of difficulty of the proposed algebraic questions. For this, we use the knowledge discovery methodology to perform the following steps from the data: cleaning, integration, selection, transformation, mining, evaluation and presentation of information. The practical results reveal that the proposed architecture can classify the academic performance of students in each evaluation period with an accuracy of around 80%. It was also possible to identify factors related to the resolution of questions, such as the rate of correct answers and the mathematical steps to solve an exercise, classifying the level of difficulty of the questions proposed by the system, acting on the student’s deficiencies, and the system. The results obtained from the data analysis can serve as a basis for decision-making, helping in the teaching process for teachers, tutors and managers to monitor academic performance, enabling the correction of problems and identifying the individual and collective difficulties present in each class. In summary, these results also provide a general roadmap on the performance of using EDM techniques in a given context.

This full paper of innovate-to-practice category presents a Multiagent System for recommending learning objects in Virtual Learning Environments (VLE), aiming to improve the customization of instructional guidance on educational content according to the student's profile. The methodology was initially research aimed at identifying the motivators of the students' performance and their weaknesses, adopting a personalized student model based on the level of knowledge, and providing predictive models to monitor the student's progress in the curriculum. This framework provides a distributed architecture, and consists of three layers: 1) Administrative layer; 2) Storage layer; 3) Pedagogical layer. For the recommendation of learning objects, a collaborative filter was used, which constitutes a successful technique in several recommendation applications, seeking similarities in users' habits to predict their future decisions.

This full paper of research category presents a proposal for a smart recommendation model to personalize the provision of teaching materials in Virtual Learning Environments (VLE), to continually learn from student feedback. This research presents a model that is divided into 4 parts: 1) Student model; 2) Domain model; 3) Mining module; and 4) Recommendation module. This research contributes to focus on the development of cognitive skills related to learning styles, aiming to help students understand the strengths and weaknesses of their cognitive and objective - cognitive strategies, with the construction of behavioral patterns generated from the exchange of messages information between the student and mining modules. It is hoped that with the implementation of e-learning models, this proposal can contribute to assist in guiding the teaching and learning process towards mastering a curriculum, with interactive regulations, and the acquisition of the corresponding skills through the discovery of knowledge and automation of the recommendation process. Besides, contributing to self-esteem and, mainly, helps to train professionals better prepared for the job market. Our focus is to describe complex processing of recommendations in e-learning environments in terms of knowledge, not the details of its implementation.

This full paper aims to present a model of the Pedagogical Agents architecture to assist in activities and tutorial interventions for mastering the resolution of first-degree polynomial equations. With the impact of the educational situation in various parts of the world due to the Covid-19 pandemic and the difficulties identified in the interaction between student and teacher, this research uses a multi-agent approach to improve the architecture of an Intelligent Tutoring System aiming to assist in the context of distance education. Exploratory research in the literature was carried out, aiming at the problem and situation, which implies the process of development of the technological artifact and from the elaboration of a product based on the gaps found. As pedagogical support, Vygotsky's social interaction theory for instructional support is used. The implementation of this proposal will result in the improvement of Tutoring Systems, allowing to guarantee conditions for the teacher and the student to carry out the teaching and learning process flexibly and intelligently, aiming at the construction of knowledge.

This full paper of research to practice category describes an experiment using an Intelligent Tutoring System to support teaching and learning that evolve solutions of first-degree equations. The difficulties encountered in solving algebra problems can make students consider math and related science a difficult task. We propose a Tutor to promote the learning of this class of problems. In this article, we describe the functioning and architecture of the tutor composed by the Interface, Student Model, Tutorial Strategies and Domain Model. Also, a Virtual Assistant was included to motivate students in the learning process. The tutoring strategies that comprise the tutor are projected from models found in the literature. To approximate the interactions of the tutor with the actions of the teachers, Lev Vygotsky's Concept of Zone Proximal Development was used. The methodology used in this project was the case study applied in a real-life context, obtaining a combination of quantitative and qualitative evidence about technical and pedagogical usability that show the degree of satisfaction of the students regarding the proposed approach.