Authors:
Dhiego Carvalho
;
Carlos Tadeu Queiroz de Moraes
;
Dante Barone
and
Leandro Krug Wives
Affiliation:
UFRGS, Brazil
Keyword(s):
Recommender Systems, Collaborative Filtering, Moodle.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence and Decision Support Systems
;
Computer-Supported Education
;
e-Learning
;
e-Learning Platforms
;
Enterprise Information Systems
;
Information Technologies Supporting Learning
;
Intelligent Tutoring Systems
;
Simulation and Modeling
;
Simulation Tools and Platforms
Abstract:
This paper proposes a recommender system for the Moodle platform that uses Apache Mahout as a recommendation engine. In this system, students receive personalized and adaptive recommendations based on forum discussions and learning materials available in the environment, according to their interactions. The recommendation method developed is based on the Moodle's activity log. In this paper, we compare four different ways to process user's preferences and present recommendations. Three of them exist in the literature, and one is proposed by us. The test set uses data from a real case scenario, i.e., the activity log from a graduation course that is given in our institution. We have found that the proposed approach produced the best set of recommendations.