Tese de Doutorado de Alessandro da Silveira Dias

Aluno: Alessandro da Silveira Dias
Orientador: Prof. Dr. Leandro Krug Wives

Título: Recommender System for Learning Objects based in Learner User Choices in E-learning Systems
Linha de Pesquisa: Planejamento, Sistemas Multiagentes e Robótica

Data: 19/02/2019
Horário: 14h30min.
Local: AUD-1 (Auditório 1) do Prédio 43412 do Instituto de Informática da UFRGS

Banca Examinadora:
– Prof. Dr. Silvio César Cazella (UFCSPA)
– Profª. Drª. Isabela Gasparini (UDESC – por videoconferência)
– Profª. Drª. Renata de Matos Galante (UFRGS)

Presidente da Banca: Prof. Dr. Leandro Krug Wives

Abstract: In this thesis, it is presented a developed recommendation approach for learning objects in e-learning systems. In these systems, learner users usually perform a set of choices or make decisions (“what to learn”, “how to learn”, “in which learning pathway to learn”, “with whom to learn”, among others) during learning, depending on the system. The developed approach uses the result of these choices as a source of information. It is an extension of the User-based Nearest Neighbor recommendation approach, which has roots in the Nearest Neighbor search problem. The result of these choices corresponds to social signals between users, and interests and preferences of learner users. With the fusion of these elements it is sought to find the most similar users to the active user, and then, to generate more accurate recommendations. An experimental evaluation of this developed approach is presented over 2 systems: the AdaptWeb and the MERLOT. It is showed, in both, that the usage prediction accuracy varies according to the combination of user choices and it presents statistically significant higher prediction than the baseline approaches. Despite being focused on e-learning systems, of the Education domain, it is discussed briefly how to use it in other domains, such as Tourism, where it is observed that users can make decisions when interacting with systems.

Keywords: Recommender System. Learning Objects. Learner-driven Learning.