Modelo I2P: recomendação de recursos baseando-se em preferências, interesses e popularidade

Abstract

The development of technologies that assist in the teach-learning process is an rgued subject in some areas of knowledge. The great diffusion of Web-based Educational Systems (WbE-S) has been shown the popularization of distance learning and its support tools. The Tidia-Ae project, support by FAPESP, aim at the development of a WbE-S that can use the concept about high velocity internet. But, the WbE Systems don t have a personal treatment of user s necessities. So, the offers of personalization resources for systems aim at improving the teach-learning process using the treatment of real necessities of each user. The content recommendation, more specifically a recommendation system, is one of several techniques for that and it is a non-intrusive meaning of help user s in a system with a lot of information. This technique was used in Tidia-Ae environment to development of this thesis. This thesis presents the I2P model based on metrics of Interests, Preferences and Popularity which are acquired by the measuring of the relationship of users and system resources. These metrics provide a form to calculate the recommendation offers of resources. The calculation is done using Collaborative Filtering technique and the personalization is offered in collaborative form, considering the group learning.