Ferramenta para Análise de Acesso de Uso de Objetos de Aprendizagem em Websites

Abstract

In this work is proposed a methodology for discovering patterns of access to Learning Objects, through data mining of web usage, ie, data is extracted from logs of websites that store all the information on how the user is related to Internet. The objective is to discover patterns of access to learning objects available on websites. The methodology used is based on the concept of KDD (Knowledge Discovery in Databases), which consists of four steps: data gathering, pre-processing, data mining and post-processing. In the first step is recorded the log file of a website that stores the learning objects. In pre-processing step was applied a filter in server logs, in order to extract only the necessary information and format them to use in the tool used for such data mining step. The third step is responsible for mining the data that were filtered and prepared in the previous step and the fourth and final step was responsible for analyzing the results generated in the mining step. Through this methodology, which can be expanded to other sites, it was possible to discover the behavior of users who make use of learning objects stored on the website.