Volume 10 - Issue 2
A Multilayer Approach for Recommending Contextual Learning Paths
- Francesco Colace
Universita degli Studi di Salerno, 84084 Fisciano SA - Italy
fcolace@unisa.it
- Massimo De Santo
Universita degli Studi di Salerno, 84084 Fisciano SA - Italy
desanto@unisa.it
- Marco Lombardi
Universita degli Studi di Salerno, 84084 Fisciano SA - Italy
malombardi@unisa.it
- Rosalba Mosca
Universita degli Studi di Salerno, 84084 Fisciano SA - Italy
rmosca@unisa.it
- Domenico Santaniello
Universita degli Studi di Salerno, 84084 Fisciano SA - Italy
dsantaniello@unisa.it
Keywords: Context-Awareness, e-learning, Recommender Systems
Abstract
Nowadays, distance learning is achieved through new technological systems, which are able to give
several advantages in the training process. Modern e-learning environments exploit technologies capable
of designing increasingly specific learning paths. This approach could be interesting in the
field of Cultural Heritage. In this scenario, the introduction of a framework able to automatically
design tailored learning paths to be used during the visit of archaeological sites could be engaging.
The proposed framework aims to exploit contextual graph approaches, such as Ontology and Context
Dimension Tree and probabilistic graph approaches such as Bayesian Networks for inferring
adapted and contextual learning paths. In particular, it supports learners during their visits in real
scenarios as archaeological parks or museums. This engine selects contents and services according
to the learner’s profile and the context. The main advantage of the proposed system is to design and
suggest tailored learning paths to be used on-site in order to improve the training process. Besides,
the proposed approach can exploit context modelling and predictive techniques, which can improve
the ability to recommend learning paths. A prototype has been developed and tested in real scenarios
as the archaeological parks of Paestum, Herculaneum and Pompeii. In particular, several aspects
have been tested, such as performance, usability and effectiveness, and more specific tests have been
performed measuring the accuracy in learning path recommendations with promising results.