Instrumento de avaliação de processos cognitivos para a aprendizagem de subestações de energia em ambientes 2D e realidade virtual

This study describes the development of a prototype to evaluate the potential of environments based on two-dimensional modeling and virtual reality as power substations learning objects into training environments from a central operation and control of power utility Cemig. Initially, there was an identification modeling features and cognitive processes in 2D and RV, from which it was possible to create frames that serve to guide the preparation of a checklist with assigning a metric weight for measuring cognitive potential learning in the study sites. From these contents twenty-four questions were prepared and each was assigned a weight that was used in the calculation of the metric; the questions were grouped into skill sets and similar cognitive processes called categories. Were then developed two distinct environments: the first, the prototype features an interactive checklist and your individual results. And, second, a system of data management environment for the configuration and editing of the prototype, and the observation and analysis of the survey results. For prototype validation, were invited to access the virtual checklist and answer it, five professionals linked to Cemig's training area. The results confirmed the validity of this instrument application to assess the possible potential of modeling in 2D and RV as learning objects in power substations, as well as provide feedback to developers of virtual environments to improve the system.

Imagens das línguas de alunos do 6º ano : um estudo em Aveiro

O presente estudo é o resultado de um projeto investigativo que, embora se relacione com a Sensibilização à Diversidade Linguística, tem como tema principal as Imagens das Línguas. Este estudo, cujo título é “Imagens das línguas de alunos do 6.º ano: um estudo em Aveiro”, tem como objetivos perceber quais as imagens das línguas dos alunos do 6.º ano, verificar qual a língua em que estes alunos se matriculam no ano seguinte e se essa escolha foi baseada em imagens estereotipadas das línguas (e em quais) e averiguar se a imagem que os alunos do 6.º ano têm sobre línguas influencia a sua escolha para aprendizagem posterior. Os dados foram recolhidos através de instrumentos distintos (o desenho e o inquérito por questionário). Primeiramente, os discentes elaboraram quatro desenhos seguindo as instruções “desenha-te a falar a tua língua materna”, “desenha-te a falar uma língua que já aprendeste”, “desenha-te a falar uma língua que gostavas de aprender” e “desenha-te a falar uma língua que não gostavas de aprender”. Seguindo-se o preenchimento do inquérito por questionário, composto por cinco questões, relacionadas com as quatro línguas em estudo (português, francês, espanhol e inglês) e que faziam parte da recolha de dados através do desenho. Relativamente ao tratamento de dados optamos pela utilização de categorias de análise (línguas como objetos afetivos, objetos de ensino-aprendizagem, instrumentos de construção e afirmação de identidades individuais e coletivas, objetos de poder e como instrumentos de construção de relações interpessoais e intergrupais), que permitiram perceber quais as imagens das línguas dos alunos inquiridos. Os resultados permitiram-nos perceber que as imagens que os alunos do 6.º ano têm das línguas portuguesa, francesa, espanhola e inglesa são, de alguma forma, estereotipadas. A maioria dos alunos tem uma imagem das línguas como instrumentos de construção e afirmação de identidades individuais e coletivas, isto é, imagens associadas à relação língua/história de um povo/cultura. Contudo, concluímos que esta imagem cultural das línguas também está associada a uma imagem afetiva, salientando a relação aluno/língua/cultura. Partindo das nossas conclusões, poder-se-ão, no futuro desenvolver sessões de Sensibilização à Diversidade Linguística, com o objetivo de (re)construir as Imagens das Línguas que os alunos têm.

Educação na era do hipertexto: uma experiência diferenciada na biologia

Acompanha: Sequência didática: trabalhando o conceito e as características dos fungos: pesquisa de campo para identificação dos fungos

Objetos de ensino, suas potencialidades e dificuldades para aprendizagem de física no ensino médio

Acompanha: Procedimento para o uso do Tracker como objeto de ensino, suas potencialidades e dificuldades para aprendizagem de física no ensino médio

Profesores frente a estudiantes: las dos orillas de la educación bimodal

Desde hace 6 años el grupo de investigación E-Virtual de laUniversidad de Medellín viene trabajando en la implementaciónde asignaturas bimodales en la Institución. En el 2009, con elapoyo de MEN, se implementó la modalidad a distancia conmetodología virtual en el modelo pedagógico de la Universidad.Estas nuevas experiencias llevaron al Grupo a cuestionarsesobre las características pedagógicas y didácticas a teneren cuenta cuando se combinan la educación presencial y lavirtual. Para ello se indagó con profesores y estudiantes sobresu percepción al respecto. Para la recolección de informaciónse combinaron técnicas cualitativas y cuantitativas, que hanarrojado interesantes resultados, entre ellos proceso deinducción, interacciones comunicativas, Objetos Virtuales deAprendizaje y uso de la plataforma virtual.En este artículo se darán a conocer algunos resultados de lainvestigación, cuáles han sido los aspectos positivos de estaexperiencia y cuáles son las áreas a mejorar.

Metodologia participativa para repositórios de recursos educacionais abertos

This research deals with the use of a participatory design methodology to develop a repository of open educational resources, the Arcaz. Discusses key aspects of neutrality and determinism of technology within the context of Social Studies of Science and Technology and presents some concepts of critical theory of technology related to the democratic construction of technological artifacts. Discusses the philosophical heritage of the movements that led to the emergence of free software, open education and open educational resources and argues that participatory design share similar ideals. It presents concepts of human-computer interaction, interaction design and user centered design, important to enhance the user experience in information systems. It addresses the participatory design as a methodology that allows the democratic participation of users in the technological construction, promoting mutual learning and active voice for the participants. Develops a participatory design methodology adapted to the Arcaz context of use and provides the procedures for the meetings conducted to apply participatory design techniques to the repository and the results obtained. It concludes with a study of some of the interventions suggested in the system and orientations for future applications of participatory practices in the development of the repository and a list of best practices, focusing on ethical principles that should guide the participatory design.

Desenvolvimento de guias didáticos para a utilização de objetos de aprendizagem

Dissertação (mestrado)—Universidade de Brasília, Instituto de Física, Programa de Pós-Graduação em Física, 2015.

Objetos de aprendizaje para el desarrollo del tema: inecuaciones de primer grado e intervalos de solución de noveno de educación general básica.

El presente trabajo tiene como objetivo dar a conocer Objetos de Aprendizaje, como un recurso tecnológico, que sirva de soporte a la educación y que contribuya en la enseñanza de las Inecuaciones de primer grado. Para realizar este proyecto se han planteado cinco capítulos de desarrollo: En el capítulo uno se realiza una investigación bibliográfica donde se describe la Educación, orientándola hacia el constructivismo, que centra al estudiante como constructor de su conocimiento y la importancia de usar material didáctico apto para desarrollar sus habilidades. En el capítulo dos se desarrolla el tema de Inecuaciones de primer grado, dividiéndolo en cuatro subtemas: Inecuaciones de primer grado con una y dos incógnitas, intervalos de solución y sistemas de inecuaciones. En el capítulo tres se presenta un análisis a los recursos educativos de libre acceso en diferentes sitios web, dónde se evidencia el desarrollo de plataformas que buscan consolidar los aprendizajes. En el capítulo cuatro se desarrolla el diseño de los cuatro Objetos de Aprendizaje; a partir de su metodología para la estructuración; además se usará guías para su organización. Y finalmente en el capítulo cinco se aprecia el interés que producen los Objetos de Aprendizaje, en los estudiantes de Noveno de EGB. Permitiendo observar que este recurso fortalece el aprendizaje de forma divertida e interactiva; dando apertura al uso de instrumentos multimedia como material de autoformación.

Bidirectional Helmholtz Machines

L'entraînement sans surveillance efficace et inférence dans les modèles génératifs profonds reste un problème difficile. Une approche assez simple, la machine de Helmholtz, consiste à entraîner du haut vers le bas un modèle génératif dirigé qui sera utilisé plus tard pour l'inférence approximative. Des résultats récents suggèrent que de meilleurs modèles génératifs peuvent être obtenus par de meilleures procédures d'inférence approximatives. Au lieu d'améliorer la procédure d'inférence, nous proposons ici un nouveau modèle, la machine de Helmholtz bidirectionnelle, qui garantit qu'on peut calculer efficacement les distributions de haut-vers-bas et de bas-vers-haut. Nous y parvenons en interprétant à les modèles haut-vers-bas et bas-vers-haut en tant que distributions d'inférence approximative, puis ensuite en définissant la distribution du modèle comme étant la moyenne géométrique de ces deux distributions. Nous dérivons une borne inférieure pour la vraisemblance de ce modèle, et nous démontrons que l'optimisation de cette borne se comporte en régulisateur. Ce régularisateur sera tel que la distance de Bhattacharyya sera minisée entre les distributions approximatives haut-vers-bas et bas-vers-haut. Cette approche produit des résultats de pointe en terme de modèles génératifs qui favorisent les réseaux significativement plus profonds. Elle permet aussi une inférence approximative amérliorée par plusieurs ordres de grandeur. De plus, nous introduisons un modèle génératif profond basé sur les modèles BiHM pour l'entraînement semi-supervisé.

Bidirectional Helmholtz Machines

L'entraînement sans surveillance efficace et inférence dans les modèles génératifs profonds reste un problème difficile. Une approche assez simple, la machine de Helmholtz, consiste à entraîner du haut vers le bas un modèle génératif dirigé qui sera utilisé plus tard pour l'inférence approximative. Des résultats récents suggèrent que de meilleurs modèles génératifs peuvent être obtenus par de meilleures procédures d'inférence approximatives. Au lieu d'améliorer la procédure d'inférence, nous proposons ici un nouveau modèle, la machine de Helmholtz bidirectionnelle, qui garantit qu'on peut calculer efficacement les distributions de haut-vers-bas et de bas-vers-haut. Nous y parvenons en interprétant à les modèles haut-vers-bas et bas-vers-haut en tant que distributions d'inférence approximative, puis ensuite en définissant la distribution du modèle comme étant la moyenne géométrique de ces deux distributions. Nous dérivons une borne inférieure pour la vraisemblance de ce modèle, et nous démontrons que l'optimisation de cette borne se comporte en régulisateur. Ce régularisateur sera tel que la distance de Bhattacharyya sera minisée entre les distributions approximatives haut-vers-bas et bas-vers-haut. Cette approche produit des résultats de pointe en terme de modèles génératifs qui favorisent les réseaux significativement plus profonds. Elle permet aussi une inférence approximative amérliorée par plusieurs ordres de grandeur. De plus, nous introduisons un modèle génératif profond basé sur les modèles BiHM pour l'entraînement semi-supervisé.

Hybrid Artificial Intelligence to Extract Patterns and Rules from Argumentative and Legal Texts

This Thesis is composed of a collection of works written in the period 2019-2022, whose aim is to find methodologies of Artificial Intelligence (AI) and Machine Learning to detect and classify patterns and rules in argumentative and legal texts. We define our approach “hybrid”, since we aimed at designing hybrid combinations of symbolic and sub-symbolic AI, involving both “top-down” structured knowledge and “bottom-up” data-driven knowledge. A first group of works is dedicated to the classification of argumentative patterns. Following the Waltonian model of argument and the related theory of Argumentation Schemes, these works focused on the detection of argumentative support and opposition, showing that argumentative evidences can be classified at fine-grained levels without resorting to highly engineered features. To show this, our methods involved not only traditional approaches such as TFIDF, but also some novel methods based on Tree Kernel algorithms. After the encouraging results of this first phase, we explored the use of a some emerging methodologies promoted by actors like Google, which have deeply changed NLP since 2018-19 — i.e., Transfer Learning and language models. These new methodologies markedly improved our previous results, providing us with best-performing NLP tools. Using Transfer Learning, we also performed a Sequence Labelling task to recognize the exact span of argumentative components (i.e., claims and premises), thus connecting portions of natural language to portions of arguments (i.e., to the logical-inferential dimension). The last part of our work was finally dedicated to the employment of Transfer Learning methods for the detection of rules and deontic modalities. In this case, we explored a hybrid approach which combines structured knowledge coming from two LegalXML formats (i.e., Akoma Ntoso and LegalRuleML) with sub-symbolic knowledge coming from pre-trained (and then fine-tuned) neural architectures.

Learning Three-Dimensional Shape Models for Sketch Recognition

Artifacts made by humans, such as items of furniture and houses, exhibit an enormous amount of variability in shape. In this paper, we concentrate on models of the shapes of objects that are made up of fixed collections of sub-parts whose dimensions and spatial arrangement exhibit variation. Our goals are: to learn these models from data and to use them for recognition. Our emphasis is on learning and recognition from three-dimensional data, to test the basic shape-modeling methodology. In this paper we also demonstrate how to use models learned in three dimensions for recognition of two-dimensional sketches of objects.

Datasets, features, learning, and models in visual recognition

Visual recognition is a fundamental research topic in computer vision. This dissertation explores datasets, features, learning, and models used for visual recognition. In order to train visual models and evaluate different recognition algorithms, this dissertation develops an approach to collect object image datasets on web pages using an analysis of text around the image and of image appearance. This method exploits established online knowledge resources (Wikipedia pages for text; Flickr and Caltech data sets for images). The resources provide rich text and object appearance information. This dissertation describes results on two datasets. The first is Berg’s collection of 10 animal categories; on this dataset, we significantly outperform previous approaches. On an additional set of 5 categories, experimental results show the effectiveness of the method. Images are represented as features for visual recognition. This dissertation introduces a text-based image feature and demonstrates that it consistently improves performance on hard object classification problems. The feature is built using an auxiliary dataset of images annotated with tags, downloaded from the Internet. Image tags are noisy. The method obtains the text features of an unannotated image from the tags of its k-nearest neighbors in this auxiliary collection. A visual classifier presented with an object viewed under novel circumstances (say, a new viewing direction) must rely on its visual examples. This text feature may not change, because the auxiliary dataset likely contains a similar picture. While the tags associated with images are noisy, they are more stable when appearance changes. The performance of this feature is tested using PASCAL VOC 2006 and 2007 datasets. This feature performs well; it consistently improves the performance of visual object classifiers, and is particularly effective when the training dataset is small. With more and more collected training data, computational cost becomes a bottleneck, especially when training sophisticated classifiers such as kernelized SVM. This dissertation proposes a fast training algorithm called Stochastic Intersection Kernel Machine (SIKMA). This proposed training method will be useful for many vision problems, as it can produce a kernel classifier that is more accurate than a linear classifier, and can be trained on tens of thousands of examples in two minutes. It processes training examples one by one in a sequence, so memory cost is no longer the bottleneck to process large scale datasets. This dissertation applies this approach to train classifiers of Flickr groups with many group training examples. The resulting Flickr group prediction scores can be used to measure image similarity between two images. Experimental results on the Corel dataset and a PASCAL VOC dataset show the learned Flickr features perform better on image matching, retrieval, and classification than conventional visual features. Visual models are usually trained to best separate positive and negative training examples. However, when recognizing a large number of object categories, there may not be enough training examples for most objects, due to the intrinsic long-tailed distribution of objects in the real world. This dissertation proposes an approach to use comparative object similarity. The key insight is that, given a set of object categories which are similar and a set of categories which are dissimilar, a good object model should respond more strongly to examples from similar categories than to examples from dissimilar categories. This dissertation develops a regularized kernel machine algorithm to use this category dependent similarity regularization. Experiments on hundreds of categories show that our method can make significant improvement for categories with few or even no positive examples.

Bayesian online algorithms for learning in discrete Hidden Markov Models

We propose and analyze two different Bayesian online algorithms for learning in discrete Hidden Markov Models and compare their performance with the already known Baldi-Chauvin Algorithm. Using the Kullback-Leibler divergence as a measure of generalization we draw learning curves in simplified situations for these algorithms and compare their performances.

Language policy spread: Learning from health and social policy models

Except for a few large scale projects, language planners have tended to talk and argue among themselves rather than to see language policy development as an inherently political process. A comparison with a social policy example, taken from the United States, suggests that it is important to understand the problem and to develop solutions in the context of the political process, as this is where decisions will ultimately be made.