Intelligent advice generator for personalized language learning through social networking sites

Computer-based advising is focusing on factors that facilitate advice giving in educational systems. In this paper, an intelligent advice generation module is presented. As a testbed for our research, we have incorporated the advice generation mechanism in a Facebook application for language learning. A significant feature in the proposed framework of advice generation is user modeling. The advice generation is required to reason about the students´ knowledge status and to decide appropriate advice. Actions, recorded by the application and transformed by the student modeler into student models, serve as the source from which diagnostic information about the student can be extracted. This information indicates which concepts were presented to the student, the duration of time the student spent working with learning objects related to a specific concept, which concepts probably mastered by the student and which concepts not yet mastered. The advice generation can identify the status of students and groups of students. This is used as a source for the generation of appropriate advice to the course instructor, who can then pass the advice to the students or consequently take some pedagogical actions that should be educationally appropriate. The advice generating mechanism based on a set of criteria for selecting appropriate advice according to the current student situation is explained in detail.