Fernando da Fonseca de Souza

GeoDWFrame: A Framework for Guiding the Design of Geographical Dimensional Schemas

Data Warehouse (DW) is a dimensional database for providing decision support by means of on-line analytical processing (OLAP) techniques. Another technology also used to provide decision support is Geographical Information System (GIS). Much research aims at integrating these technologies, but there are still some open questions, particularly regarding the design of a geographical dimensional data schema. This paper discusses some related work and proposes GeoDWFrame that is a framework based on the star schema and has been specified as guidance for designing geographical dimensional schemas. Some experimental results are also given.

Horizontal fragmentation as a technique to improve the performance of drill-down and roll-up queries

In this paper, we focus on the horizontal fragmentation of data warehouses. Our main contribution is the proposal of the MHF-DHA algorithm, which is aimed at improving the performance of drill-down and roll-up queries by horizontally fragmenting data warehouses organized in different levels of aggregation. Besides allowing that multiple dimensions be used as a basis for the fragmentation, the algorithm also explores the hierarchical structure of these dimensions. The performance tests carried out using the TPC-H benchmark showed that the proposed fragmentation provides a huge improvement on the query performance, with a reduction in elapsed time and disk accesses between 71% and 99%.

Focusing on Data Distribution in the WebD2W System

The WebD2W system is a distributed client-server data warehousing environment, which is aimed not only at the data warehouse distribution, but also at the distributed access to these data using the Web technology as an infrastructure. In this paper, we introduce the WebD2W system, focusing on one of its main objectives: the data warehouse distribution. Such a system is presented in terms of its main components and their respective functionalities. The paper also describes the algorithm for fragmenting horizontally the warehouse data, which is used as a basis for the WebD2W system.

An intelligent Web interface to generate and update adaptive Virtual Environments

In recent years, three-dimensional Virtual Environments developed with virtual reality and graphical computation technologies have evolved in the Web. Thus, users with different profiles, cultures and knowledge levels can access such environments remotely. In such a way, it is necessary to use an efficient interface to manage Web accesses and to actualize virtual environments to adapt them to each users characteristics. This article presents an intelligent Web interface managed by a society of software agents that follows and identifies users actions in order to propose modifications in the virtual environment, through 3D objects updating. The interface is capable to identify users actions and to react to them in real time, without loss of performance. The objects used in the construction and adaptation of a Virtual Environment are stored into a database and are retrieved by agents, according to the specific negotiation rules defined in an ontology for the domain.

Discovering Level of Participation in MOOCs through Clusters Analysis

This paper presents an analysis using hierarchical grouping method (ward grouping) and the non-hierarchical grouping method (k-means) to analyze the participation levels in activities and interactions in a virtual forum. Data came from a MOOC and it was on grammatical rules of Brazilian Portuguese. About 5100 participants integrated the course. It lasted about three months and the activities developed by means of the Openredu learning platform. We analyzed data of collaboration, interaction, and discussions in forums, access data and activity on the Openredu. The results pointed to three distinct engagement strategies. Those categories oriented the proposition of an interface design guidelines for MOOC to conceive adaptive strategies that permit to increase engagement and favor an improved learning experience.

Adaptive Virtual Environments: The Role of Intelligent Agents

A virtual environment (VE) can simulate a real environment or provide imaginary or even physically impossible scenarios. It allows one or more users to interact through visualization and manipulation of extremely complex representations (Frery, 2002). According to Chittaro & Ranon (2007), a VE can help people through direct experience by performing tasks that are suited to the learning or simulation tasks. Moreover, employing interactive 3D objects allows for more useful representations of subjects or sensations, offering the possibility of analyzing the same subject from different points of view. VE is defined here as a computer-generated, graphically-rich 3D world which the user can explore and perform actions on real time. The environment is based on 3D computer models and the user can navigate on the environment and interact with the 3D objects and other avatars existing there. The user’s view of the environment is rendered dynamically and can be updated as the user moves.

Mineração de Dados Educacionais Orientada por Atividades de Aprendizagem

Este trabalho, por meio de um estudo de caso, apresenta uma abordagem de Mineracao de Dados Educacionais orientada por atividades de aprendizagem, tendo como referencia a Teoria da Atividade. O objetivo do estudo foi analisar as diferencas nos resultados dessa abordagem em relacao a um processo de mineracao holistico, no qual os modelos de predicao permitem analises apenas no nivel de disciplina, sem a observacao de detalhes das atividades de aprendizagem. Os resultados desta pesquisa apontam vantagens da mineracao orientada por atividade, que oferece informacoes em um contexto de interacao significativa, com mais subsidios para monitorar e tratar contradicoes no processo de aprendizagem.

Mapeamento Sistemático sobre Abordagens de Mensuração de Autorregulação da Aprendizagem

Este trabalho teve como objetivo o desenvolvimento de um mapeamento sistematico com a finalidade de identificar as principais abordagens de mensuracao de caracteristicas de autorregulacao da aprendizagem em ambientes online. A identificacao destas abordagens fornecem subsidios para identificar as principais formas de analises e coletas que estao sendo utilizadas na literatura. Os resultados mostram que ainda e pouca expressiva a quantidade de trabalhos que utilizam dados comportamentais, atraves de logs, para a mensuracao em tempo real. Este parece ser um tema desafiador especialmente considerando o numero cada vez maior de cursos a distância, numero de alunos e volume de dados gerados em plataformas de aprendizagem online.

Análise do engajamento de estudantes com base na Distância Transacional a partir da Mineração de Dados Educacionais

Este estudo apresenta uma proposta de analise do engajamento de estudantes em um curso online. Foram usados dados de um curso de pos-graduacao de uma universidade publica brasileira. O metodo seguiu o processo de Mineracao de Dados Educacionais aplicada em dois ciclos, a partir da identificacao dos construtos da distância transacional nos dados coletados, da utilizacao de metricas definidas pela analise de redes sociais e da aplicacao de regressao logistica para obtencao de modelos representativos do engajamento e dos fatores que o influenciam. Os resultados apontaram caminhos para intervencoes que poderiam ser feitas para o aumento no nivel de engajamento dos estudantes.

Validação de um instrumento de mensuração de autorregulação da aprendizagem em contexto brasileiro usando análise fatorial confirmatória

A escala de autorregulacao da aprendizagem Online Self-Regulated Learning Questionnaire (OSLQ), desenvolvida por (Barnard, Lan e To, et al. 2009), e um instrumento para avaliar as caracteristicas de autorregulacao da aprendizagem de estudantes em cursos na modalidade online. O presente estudo visa investigar a aplicabilidade deste instrumento, verificando se o mesmo e valido para mensurar caracteristicas de autorregulacao da aprendizagem nos moldes da educacao a distância no Brasil. Foi realizada uma pesquisa com 408 participantes de cursos na modalidade EAD com idade media de 30 anos (DP=18,23). A analise dos dados foi realizada por meio de analise fatorial confirmatoria e os resultados indicaram que o modelo de seis fatores mensurado atraves dos itens do instrumento se ajusta a amostra analisada no Brasil. Consideracoes acerca da validade interna e da proposta conceitual do instrumento sao discutidas neste artigo.