This study investigates sensor-free emotion detection techniques in Intelligent Learning Environments, focusing specifically on detecting engagement and confusion. Traditional sensor-based methods of emotion detection face challenges, especially in low- and middle-income countries due to resource constraints. Consequently, our research utilizes data mining and machine learning algorithms to develop models that accurately detect student emotions using only interaction data. We also examine the potential of incorporating personality traits to boost the effectiveness of these models. Our analysis involves a dataset of interactions and annotated emotions from students using a Math step-based Intelligent Tutoring System. We implemented a structured machine learning pipeline designed to prevent data leakage and ensure robust model development. The findings indicate that while the baseline models for detecting engagement and confusion are reliable and show improvements over previous models tested on the same dataset, the inclusion of personality traits did not significantly enhance model performance. This research highlights the complexity of emotion detection in learning environments and discusses the implications of using personal characteristics to improve emotion detection accuracy.

Misconceptions play a significant role in the learning process as they reflect an inaccurate understanding of a particular concept. Error diagnosis can help teachers and intelligent learning environments determine the most appropriate type of student assistance. Previously, misconceptions were identified using rule-based expert systems (bug libraries) and clustering algorithms. Bug libraries demand extensive work from developers to identify all potential misconceptions and code rules for each one in advance. Additionally, these solutions cannot detect misconceptions for which rules were not explicitly programmed. Clustering-based solutions overcome these drawbacks by automatically identifying misconceptions based on students' most common errors. To effectively and efficiently identify misconceptions, clustering solutions must have a suitable representation of the problem and its steps, and employ machine learning algorithms capable of discerning patterns from them. This paper proposes a solution that utilizes expression trees to represent algebraic problem-solving steps and the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to identify misconceptions by clustering similar errors in a database containing 1064 steps from 112 students. This database was collected from an intelligent learning system designed to assist in solving first-degree equations. In our final solution, a Natural Language Processing tokenizer was employed to represent each term numerically, which identified 178 homogeneous clusters with minimal noise and few outliers.

The gamification of education can help generate levels of involvement and dedication similar to games, enhancing specific skills, engaging students, and maximizing learning. However, some studies have shown detrimental results depending on the user's characteristics. It is still necessary to study the relationship between the user's profile, activity performed and the gamification elements applied. With this goal, this work incorporated gamification with points, ranking and badges into a programming learning environment to study the effects of these gamification elements on students' learning, behavior and en-gagement according to their personality trait of extroversion. An experimental evaluation was carried out during four months with 40 undergraduates students of first-year introductory courses on programming, in which half of the students used the gamified version of the learning environment and the other half the same learning environment with no gamification. We have found evidence that gamification affected the engagement of introverted and extroverted participants differently. Introverted students who used the gamified version of the educational learning system submitted a greater number of correct solutions, had shown to be more engaged to gain points and badges, and had highest ranking visualization when compared to extroverts of the gamified group. These results contribute to the understanding that the effect of gamification depends on users' specific characteristics, such as the personality trait.

Intelligent Tutoring Systems (ITSs) for Math still use traditional methods of input, such as computer keyboard and mouse. This strategy exposes students to a high extraneous cognitive load, due to the way the students are used to work, what could impair their learning. This paper presents a tool for Math ITSs that uses the student handwriting as data input. This tool was integrated into PAT2Math, an algebraic ITS, for an experimental evaluation in a public school with 48 students. It followed a design of control and experimental groups with pre and post-tests and a post questionnaire. Although our results have shown statistically significant improvement in student learning gains in both groups, no difference between groups was found. However, questionnaire responses suggest that handwriting provides a greater facility to the insertion of equations into ITSs for Math if compared with traditional methods (i.e., keyboard and mouse).

This article presents a logical formalization of the emotional appraisal theory, i.e., it formalizes the cognitive process of evaluation that elicits an emotion. This formalization is psychologically grounded on the OCC cognitive model of emotions. More specifically, we are interested in event-based emotions, i.e., emotions that are elicited by the evaluation of the consequences of an event that either happened or will happen. The formal modelling presented here is based on the AfPL Probabilistic Logic, a BDI-like probabilistic modal logic, which allows our model to verify whether the variables that determine the elicitation of emotions achieved the necessary threshold or not. The proposed logical formalization aims at addressing how the emotions are elicited by the agent cognitive mental states (desires, beliefs and intentions), and how to represent the intensity of the emotions. These are important initial points in the investigation of the dynamic interaction among emotions and other mental states.

Animated Pedagogical Agents (APAs) may be used to help students in the process of learning a foreign language in Computer Assisted Language Learning (CALL) by, for instance, pointing at specific areas of content to direct attention. Previous work has analyzed APAs in CALL systems, but results were contradictory and inconclusive regarding to whether APAs help students perform better or not. Additionally, none of the studies we have found studied APAs in CALL systems lecturing about the English present perfect tense, which is considered one of the main issues Brazilian students face when learning English as a foreign language. This paper describes the design, implementation and analysis of an APA integrated to a CALL system to help Brazilian students of English as a foreign language learn the present perfect tense. Thirty-nine Brazilian undergraduate students were randomly assigned to control and experimental groups, to evaluate whether the addition of an APA with features like talking, moving and pointing helps students improve performance. Pretest and posttest results were evaluated using statistical tests, which showed a significant increase in performance of students who interacted with the agent, when compared to correspondent outcomes of students in the control group.

Infographic is a type of information visualization that uses graphic design to enhance human ability to identify patterns and trends. It is popularly used to support spread of information. Yet, there are few studies that investigate how infographics affect learning and how individual factors, such as learning styles and enjoyment of the information affect infographics perception. In this sense, this paper describes a case study performed in an online platform where 27 undergraduate students were randomly assigned to view infographics (n=14) and graphics+text (n=13) as learning materials about the same content. They also responded to questionnaires of enjoyment and learning styles. Our findings indicate that there is no correlation between learning styles and post-test scores. Furthermore, we did not find any difference regarding learning between students using graphics or infographics. Nevertheless, for learners using infographics, we found a significant and positive correlation between correct answers and the positive self-assessment of enjoyment/pleasure. We also identified that students who used infographics keep their acquired information longer than students who only used graphics+text, indicating that infographics can better support robust learning.

This paper presents a dynamic Bayesian network model for the assessment of students' algebraic knowledge in step-based intelligent tutoring systems. The proposed work assesses knowledge about concept, skills, and misconceptions of learners. Furthermore, the proposed model is independent of the problems provided by the system (i.e., equations), because it considers the algebraic operation used by the student to solve a step as evidence instead of the final solution provided by the student. Results of evaluations comparing student's performance on a posttest with the inference of the proposed model showed statistically significant similarities between them, indicating that the inference performed by the model was accurate.

Affective states such as personality trait, emotion and mood are critical elements that affect group learning. Nevertheless, few studies classify and analyze the research findings of the community to explicitly show the real impact and use of affective states in CSCL environments. Thus, this paper presents the development of a systematic mapping of the literature to answer three questions: (i) What are the affective states frequently used in CSCL environments? (ii) Were affective states in CSCL environment empirically evaluated? (iii) What are the types of environments that use affective states to support group learning? To answer these questions, seven digital libraries were queried and 31 studies were deeply analyzed. As main results, we verified that: (i) 54,84% of the studies deal with 'emotions' in CSCL environments, (ii) 51,61% of the studies were empirically evaluated, and (iii) LMS (38,71%) and E-Learning systems (22,58%) are the most common types of environments that use affective states to support collaborative learning. Finally, we also identified gaps and opportunities to conduct research on adaptive/intelligent CSCL systems to support better group formation and customization of group activities using affective states.

A very well known reasoning model in AI is the BDI (Belief-Desire-Intention). A BDI agent should be able to choose the more rational action to be done with bounded resources and incomplete knowledge in an acceptable time. Although humans need emotions in order to make immediate decisions with incomplete information, traditional BDI models do not take into account affective states of the agent. In this paper we present an implementation of the appraisal process of emotions in BDI agents using a BDI language that integrates logic and probabilist reasoning. Specifically, we implement the event-generated emotions with consequences for self based on the OCC cognitive psychological theory of emotions. We also present an illustrative scenario and its implementation. One original aspect of this work is that we implement the emotions intensity using a probabilistic extension of a BDI language. This intensity is defined by the desirability central value, as pointed by the OCC model. In this way, our implementation of emotional BDI allows to differentiate between emotions and affective reactions. This is an important aspect because emotions tend to generate stronger response. Besides, the intensity of the emotion also determines the intensity of an individual reaction.

The interface is the main mechanism of communication between user and system features. In educational software, successful user interface designs minimize the cognitive load on users, thereby users can direct their efforts to maximize their understanding of the educational concepts being presented. We investigated whether a reduced interface make few cognitive demands on users in comparison to a complete interface. In this context, this research aims at analyzing a reduced and a complete interface of an interactive geometry software, and verify the educational benefits they provide. To this end, we designed the interfaces and carried out an experiment involving 69 undergraduate students. The experimental results indicate that an interface that hides advanced and extraneous features helps novice users to perform slightly better than novice users using a complete interface. After receiving proper training, however, a complete interface makes users more productive than a reduced interface.

Nowadays, socio-affective factors such as cooperation, motivation, socio-cognitive affinities, pro-activity, interaction and others, tend to be widely exploited in teacher/student and in between pupils relationship, so that the results of the building of knowledge be reached in a collaborative fashion. This is because these factors would play an encouraging role towards collaboration, acting positively in building human relationships within a socio-cultural panorama. Therefore, this work will show the importance of considering socio-affective factors in collaborative learning environments focusing on the relevant socio-affective factor that we designate as Collaboration Capacity, describing its manual and computed aided inference mechanisms. The rest of the relevant socio-affective factors quoted in this work, will be described and inferred along the course of the dissertation concluding research.