A Proposed Ontology-Based Generic Context Model for Ubiquitous Learning

Introduction

Ubiquitous learning (u-learning) can be seen as a new paradigm regarding deliver adaptation of learning services for learners. It aims to provide learners with the appropriate information, at the appropriate time, in the appropriate way, using mobile devices, wireless connection, and sensor technology (González et al., 2016). One essential feature to realize this type of smart learning is context awareness. A context-aware system is able to collect characteristics of the environment that surround the learner and respond to different changes (González et al., 2016). There has been a growing research background about the topic of context modeling and several approaches have been proposed (Yin et al., 2015), out of which ontology seem to be the most successful and promising one, this is due to the fact that it construct expressive models that can manage the required context information and offer some basic reasoning mechanisms. Another important purpose of ontology consists of providing a common understanding vocabulary for a specific field. For, it might be advantageous to have a generic ontology-based context model to unify contextual information in the learning environment, however, according to several studies (Economide, 2009; González et al., 2016; Ennouamani & Mahani, 2018; Yin et al., 2015), a large amount of works in the area of applying ontology within the e-learning field, address the personalization of learning objects based only on a subset of learners’ context characteristics that is of interest in a particular learning scenario (Soualah-Alila et al., 2013; Naqvi et al., 2013; Salazar, 2014; Yin et al., 2015, Abech et al., 2016; González, 2014; Popović et al., 2016; Sevkli et al., 2017; Gómez et al., 2016; Skoulikari et al. 2015, Marcelo et al., 2016, Louhab et al., 2018, Aeiad & Meziane, 2018). As such, there is no proposal of a generic context model for the context-aware u-Learning.

The lack of this kind of models is due to its complexity because it is considered a model, which has to handle all situations (Economide, 2009), in this case, the accuracy of the application based on this context is increased in parallel with its complexity and the requirements to collect data. Regarding (Economide, 2009, yin et al., 2015), there should be a balance between the number of contextual parameters, model complexity, and model accuracy, for that reason, it is necessary to reduce the context model’s complexity by extracting useful context information.

In this article, the authors try to answer the following problem statement: which characteristics should be integrated into the proposed generic context model? In fact, the identification of the characteristics which effectively influence learning and which meet the learners’ requirements becomes the most difficult end to achieve. This is because of the large number of context parameters. For that purpose, our research aims to contribute in the field of adaptive mobile learning by introducing a generic model for learners’ context including the most common and frequently used characteristics within this field (Aguilar et al., 2018, Yin et al., 2015).

The reminder of this paper is articulated as follows. Section 2 analyzes several existing context models in u-Learning field, in order to conclude which characteristics should be involved in the model. Then the architecture of the proposed system is provided in section 3, it aims to show the position of the proposed generic Context model in the future system. Section 4 presents in details the proposed context model. Section 5, demonstrates the practicality and the richness of the proposed ontology through several scenarios and a comparative study. The conclusion of this work is introduced in Section 6.