Visualização de informações para acompanhamento de alunos em ambientes de aprendizagem eletrônica

• Autor(a) principal: Macedo, Maylon Pires
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/12818
• topic: Infovis; Dados Educacionais; EDM; Educational Data; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO

The Information Visualization area studies ways to present data graphically in an efficient way to assist the user in decision making. Education professionals have been increasingly interested in the inclusion of visualization tools to improve the analysis of Educational Data. However, tools for delivering visualizations on Educational Data do not often provide resources aligned to the needs of the target audience. The objective of this master's project was to propose a guide, called Vis2Learning, composed of recommendations to help the developer build visualizations about Educational Data applied to the context of e-learning. For each application scenario covered by the guide, the following information are provided: the most suitable graph format; your characteristics; application examples; and general recommendations to bring improvements to the data visualized through the graph. Based on a preliminary study of the literature, were found some gaps related to the process of developing visualizations in the educational scenario. A Systematic Literature Review (RSL) was conducted to understand what approaches were used in this context. Based on the results of RSL, was discovered that when developing visualizations on Educational Data, researchers tend to perform exploratory research and develop visualizations based only on examples. Vis2Learning was proposed based on an analysis of studies and results reported in the articles obtained in the initial stages of RSL. From this analysis, lessons learned during the evaluation of the visualizations developed for the educational scenario were extracted. These extractions were validated by a trio of specialists in: Educational Data Mining; Human-Computer Interaction; and Information Visualization. After extracting and validating the recommendations, the final version of Vis2Learning was defined. Vis2Learning was evaluated from two perspectives, namely: an assessment aimed to collect the perception of end users about the adequacy of views developed based on Vis2Learning for the educational scenario; while in the other evaluation, was analyzed the utility and ease perceived by developers in the use of Vis2Learning for the development of visualization prototypes on Educational Data. The result of the evaluation with the end users showed that traditional views built based on Vis2Learning were considered more suitable for the educational scenario in relation to the views built by the Google Sheets algorithm. In the evaluation with developers, the Vis2Learning helped participants to gain awareness of good practices in the development of visualizations and supported the process of choosing the visualization format taking into account the user's context.