Search

An intelligent tutoring system is used as an efficient self-learning tutor, where decisions are based on the affective state of the user. These detected emotions are what experts call basic emotions and the best-known recognition technique is the recognition of facial expressions. A convolutional neural network (CNN) can be used to identify emotions through facial gestures with very high precision. One problem with convolutional networks, however, is the high number of hyperparameters to define, which can range from a hundred to a thousand. This problem is usually solved by an expert experience combined with trial and error optimization. In this work, we propose a methodology using genetic algorithms for the optimization of hyperparameters of a CNN, used to identify the affective state of a person. In addition, we present the optimized network embedded into an intelligent tutoring system running on a mobile phone. The training process of the CNN was carried out on a PC with a GPU and the trained neural network was embedded into a mobile environment. The results show an improvement of 8% (from 74 to 82%) with genetic algorithms compared to a previous work that utilized a trial and error method.

Pattern recognition is an important skill of Computational Thinking and is one of the most important competences for solving a problem that involves finding similarities or patterns in small problems to solve more complex ones. In this work, we present the mobile application software Patrony. The main contribution of this work is to promote the learning of Computational Thinking, especially pattern recognition, in specific sectors of education in Mexico through the simple use of a software application. To evaluate the effectiveness of the mobile application, tests were carried out in two elementary schools with a total of 43 students, which were divided into 2 groups: a control group and an experimental group. The results of the tests showed that the learning gain (M = 6.50 in postest compared to M = 4.94 on pretest) of the students who used our mobile application produces a significant difference with respect to students who learned using a traditional method of classroom teaching. The results also infer that computational thinking applications can be used as effective learning tools within some important Mathematics topics in public and private schools in Mexico.

This paper presents a comparison among several sentiment analysis classifiers using three different techniques – machine learning, deep learning, and an evolutionary approach called EvoMSA – for the classification of educational opinions in an Intelligent Learning Environment called ILE-Java. To make this comparison, we develop two corpora of expressions into the programming languages domain, which reflect the emotional state of students regarding teachers, exams, homework, and academic projects, among others. A corpus called sentiTEXT has polarity (positive and negative) labels, while a corpus called eduSERE has positive and negative learning-centered emotions (engaged, excited, bored, and frustrated) labels. From the experiments carried out with the three techniques, we conclude that the evolutionary algorithm (EvoMSA) generated the best results with an accuracy of 93% for the corpus sentiTEXT, and 84% for the corpus eduSERE.

In this paper we present the creation process of a corpus of phrases (opinions) related to learning computer programming. Opinions (textual phrases) are categorized in different emotions related to learning such as frustration, boring, excitement, and engagement. The results shows that in 851 opinions (754 Positives and 97 Negatives), there was a neutral, excited and engaged emotional tendency which indicates that we must include resources that induce students to write negative comments. We tested the corpus with a set of machine learning classifiers where the classifier with the highest score was Bernoulli Naive Bayes with an accuracy of 76.77%.

This paper presents the design and implementation of an intelligent tutoring system (ITS) for teaching JAVA, which can recognize the user's emotional state through facial expressions and textual dialogues. For facial emotion recognition we implemented a neural network with WEKA library and a facial feature extractor with OPENCV library. The ITS applies a semantic algorithm (ASEM) to extract textual emotions through dialogues, which has shown a degree of assertiveness of 80% in tests for graduate students. In addition, the tutor uses a set of fuzzy rules to determine the complexity of the next exercise, considering the program implementation time, program executions and compilations, and current difficulty level.

We have developed an affective and intelligent learning environment that helps students to improve their Java programming skills. This environment evaluates cognitive and affective aspects of students in order to define the level of difficulty of the exercises that are more suitable for the them in its current condition. The cognitive aspects are: the number of mistakes, the difficulty level of the current exercise and the time spent in the solution. The affective aspects are: the acquired emotion from a facial expression and the acquired valence from electroencephalogram signals. This environment also uses a neural network for face recognition of basic emotions, a support vector machine to define the valence of emotion and a fuzzy inference engine to evaluate the cognitive and affective aspects.

Learning-oriented emotions have not been studied by emotion recognition systems. These emotions have not been taken into account by other studies despite their importance in educational context. This work presents a recognition system which uses deep learning approach using convolutional neural network for solving that problem. A convolutional architecture was designed and tested with 3 different facial expression databases. The architecture is composed of 3 convolutional layers, 3 max-pooling layers, and 3 deep neural networks. The first database contains facial images on 6 basic emotions; the second and third databases contain images of learning-centered facial expressions. The tests show a 95% in the basic emotion database, a 97% for the first learning-centered emotion database and a 75% for the third database. We discuss about the differences in results among the three emotion databases.

This work presents the building and validating of a face expression database and a face expression recognizer. The face expression recognizer uses a geometric-based technique that measures distances between the central point on the face and other 68 facial landmark points. These measures are transformed into features to train a support vector machine. The database was built inside an educational context while students program in Java code. The tests validate the accuracy of the recognizer applying a ten-fold cross-validation.

Studies investigating the effectiveness of affect detection inside intelligent learning environments (ILEs) have reported the effectiveness of including emotion identification on learning . However, there is limited research on detecting and using learning-centered data to investigate metacognitive and affective monitoring with ILEs. In this work we report the methodology we follow to create a new facial expression corpus from electroencephalography information, an implementation of an algorithm and a training of an SVM to recognize learning-centered emotions (frustration, boredom, engagement and excitement). Also, we explain changes realized in a fuzzy logic system into an intelligent learning environment. The affect recognizer was tested into an ILE for learning Java programming. We present successful results of the recognizer using our corpus face database and an example test using our ILE.

The growing demand for software tools that encourage and support students in learning design and algorithm implementation, has allowed the creation of such software systems. In this paper we present a new and innovative affective tutoring system, for logic and algorithmic programming, based on block techniques. Our approach combines the Google Blockly’s interface with gamification techniques and exercises that are monitored according to the emotional state of the student. Depending on the expressed emotion (boring, engagement, frustration, and neutral), the system evaluates a number of variables to determine whether the student requires assistance. Tests have shown that the detection of the emotional state of the student, affect favorably the student evaluations.

We present Fermat, an Intelligent Social Network for Mathematics Learning, which integrates an Intelligent Tutoring System as an extra feature to help improve the teaching and learning process. The intelligent tutor takes into account both cognitive and affective factors, and by use of artificial intelligence techniques, provides users with a personalized and satisfying experience when taking the courses.

This paper presents a Web 2.0 Learning Environment, for a systematic creation of adaptive and intelligent tutoring systems. Authoring contents is made by a community of users including teachers and students. The tutoring systems adapt the contents according to the best learning style using self-organizing maps (SOMs). The SOM was trained for classifying Felder-Silverman learning styles. The most important advantage of these unsupervised neural networks is that they do not require an external teacher for presenting a training set. The approach was implemented under an authoring tool that allows the production of personalized learning material to be used under collaborative and mobile learning environments. The tutoring systems together with the neural network can also be exported to mobile devices. We present different results to the approach working under the authoring tool.

This paper presents two software systems for visual affect and learning styles recognition. The first system recognizes Paul Ekman's seven basic emotions in student expressions which are surprise, fear, disgust, anger, happiness, sadness, and neutral. The second system recognizes the student learning style using the Felder-Silverman Model. Both systems are integrated into an intelligent tutoring system in a math social network. The automatic recognition was implemented using Kohonen networks which were trained to recognize and classify emotions and learning styles. We show and discuss results by using different methods with respect to affect or emotion recognition and present the automatic response to affect results. We also present the software architecture where both recognizers collaborate with intelligent tutoring systems in a social network.

In this paper we present step by step the design and implementation of affective tutoring systems inside a learning social network using soft computing technologies. We have designed a new architecture for an entire system that includes a new social network with an educational approach, and a set of intelligent tutoring systems for mathematics learning which analyze and evaluate cognitive and affective aspects of the learners. Moreover, our intelligent tutoring systems were developed based on different theories, concepts and technologies such as Knowledge Space Theory for the domain module, an overlay model for the student module, ACT-R Theory of Cognition and fuzzy logic for the tutoring module, Kohonen neural networks for emotion recognition and decision theory to help students achieve positive affective states. We present preliminary results with one group of students using the software system.

Integrating teaching with students' emotions is seeking to optimize the learning of these students. This paper presents a learning system which combines different technologies like a learning social network or knowledge society, an authoring tool to produce intelligent tutoring systems and a system for emotion recognition. The recognition of affection or emotions is through modular neural networks which integrate the results of recognizing emotions in faces and voice. The system that recognizes emotions is integrated with the intelligent tutoring systems which in turn are integrated into the learning social network.

Teaching a course in compiler construction is considered always a challenge because there are several problems to be addressed as time, complexity and motivation of students. In this paper, we present a hybrid learning approach along with a tool for use with courses of compiler construction. The key to our method is to combine theoretical and practical topics of the course using various technologies such as mobile learning, intelligent tutoring systems, learning social networks with direct learning. The ultimate goal is to stimulate the student's abilities to work creatively, collaboratively or individually, as well as their ability to solve complex problems.

Automatic Infant Cry Recognition (AICR) process is basically a problem of pattern processing, very similar to the Automatic Speech Recognition (ASR) process (Huang, Acero, Hon, 2001). In AICR first we perform acoustical analysis, where the crying signal is analyzed to extract the more important acoustical features, like; LPC, MFCC, etc. (Cano, Escobedo and Coello, 1999). The obtained characteristics are represented by feature vectors, and each vector represents a pattern. These patterns are then classified in their corresponding pathology (Ekkel, 2002). In the reported case we are automatically classifying cries from normal, deaf and asphyxiating infants. We use a genetic algorithm to find several optimal parameters needed by the Fuzzy Relational Neural Network FRNN (Reyes, 1994), like; the number of linguistic properties, the type of membership function, the method to calculate the output and the learning rate. The whole model has been tested on several data sets for infant cry classification. The process, as well as some results, is described.

L2Code is an Intelligent Tutoring System used for teaching programming courses for different paradigms under a hybrid or blinded environment. It was designed and implemented to work with diverse types of modules oriented to certain ways of learning using principles of Multiple Intelligences. The author tool facilitates the creation of adaptive or personalized learning material to be used in multiple-paradigm programming language courses applying an artificial intelligence approach. The Tutoring System works with a predictive engine that uses a Naive Bayes classifier which operates in real time with the knowledge of the historical performance of the student. We show results of the tool.