Introduction
E-learning has allowed courses spread through the world overcoming time and space boundaries established in classrooms. This produced many benefits: enhanced student participation, better communication between students and teachers, and motivation to work in a collaborative way, among others. However, all technologies change over time and this context is not an exception. It is necessary not only innovate, but also solve problems and challenges that arise day to day.
The application of information and communication technologies (ICT) in education does not mean that presence of a teacher who plans, organizes, guides and assesses learning should be left aside. ICT must assist teachers in developing better pedagogical and didactical methods taking into account differences between students. Some studies such as those introduced by Marques, Villate, and Carvalho (
2017); Meira Ferrão Luis, Llamas-Nistal, and Fernández Iglesias (
2017); Morais, Alves, and Miranda (
2017) highlight the importance of analysing data generated by Learning Management Systems (LMS). These systems gather a lot of information about student’s action and activities, but teachers have difficulties to analyse and use it in order to improve their strategies.
Another aspect to take into account is the assessment of learning. Some researches evidence the importance of this process since it influences on the learning quality. So, student learning depends on how they are assessed and with which instruments (García Carreño,
2012). Moreover, assessment is a better way to keep students interested because they feel observed, guided and controlled. So, teacher assesses student to identify the evolution through different learning levels. On the one hand, according to the level achieved, instructor proposes actions and materials to improve student performance (Leyva Leyva, Proenza Garrido, Leyva Leyva, Varona, & Romero Rodríguez,
2008) such as assessments, videos, essays, lectures, among others. On the other hand, each student learns at his/her own pace. This creates a need to personalize the mechanisms used according each student. Technologies can help teacher to discover who is learning and who is unmotivated. They may propose actions like when, whom and how assess or what activities can be done.
In this pedagogic-technological context, self-regulation of learning becomes a key factor. It is necessary that students use the technologies to plan, organize and facilitate their own learning. One of the available technologies to support this process is the e-Portfolio. An e-Portfolio is a digital folder where students collect and include material that evidences their learning progress. This helps them to identify which aspects dominate and which ones should improve (Barberà, Bautista, Espasa, & Guasch,
2006). In addition, it allows teacher to better understand the individual characteristics of each student and attend different learning paces that can coexist in their group (Rico Martín,
2010). The e-portfolio must be prepared to support different digital materials, activities and learning methodologies adopted by teachers that can change over time. Using semantic technologies in its development is a step forward in order to support personalization and self-regulated learning.
According with related works, different aspects must be taken into account: (i) the use of learning path and learning levels in e-learning, (ii) the use of e-portfolios in the learning process and how it facilitates self-regulation learning and (iii) the use of semantic technologies to develop this tool.
With respect to the first one, Peres, Oliveira, Jesus, and Silva (
2017) present an implementation of learning path into e-learning platform, taking into account the seven learning level defined by Bloom (Bloom, Engelhart, Furst, Hill, & Krathwohl,
1956). However, it does not allow students to take a different path according with their capabilities and learning pace.
As regards the use of e-portfolios in the learning process, Encalada, Santiesteban, Portela, Cruz, and Arboleda (
2017) implement an e-portfolio in the Mahara platform adapted to the Lesson Study model, which facilitates the collaborative work of the learners in Educational Sciences careers in their pre-professional practices. Weber and Myrick (
2018) analyse the experiences of undergraduate students in a summer research program and the development of e-portfolios through free web pages such as Wix and WordPress. Chittum (
2018) suggests, in an introductory course of educational psychology for teachers, the use of deepen learning and motivation thought the implementation of e-Portfolios based on the
Digication web platform. Collins and O’Brien (
2018) use e-portfolio implemented in the
Pathbrite platform for the nursing bachelor program of the School of Nursing, at Otago Polytechnic. Many of these underline aspects correspond to what Alexiou and Paraskeva (
2010) mention: the use of e-portfolios implies carrying out processes of reflection, planning, synthesis, presentation, discussion and feedback, which are part of self-regulated learning.
In addition, the above-mentioned works have other aspect in common. They do not use semantic technologies in the implementation of the e-portfolio. They use web pages or platforms available on the Internet contributing to the lack of personalization and interoperability. There are specifications such as IMS ePortfolio and JISC Leap2A that can be used for developing new e-portfolio systems solving interoperability problems. Nevertheless, they are based on the XML language, in which the semantic of the elements is comprehensible by humans but not by machines (Rezgui, Mhiri, & Ghédira,
2018).
Finally, in respect of use of semantic web technologies in e-portfolio implementation, Taibi, Gentile, Fulantelli, and Allegra (
2010) develop an ontology that reuses and extends the concepts of the FOAF (Friend Of A Friend) ontology to include e-portfolio concepts, but it is oriented only to informal learning activities. Wang and Wang (
2012) describe an ontological approach to organize the resources in an e-portfolio, but it is not implemented in a formal language like OWL. Nguyen and Ikeda (
2014) propose an ontological model that is used to design and implement an ePortfolio system in order to promote self-regulated learning. Rezgui et al. (
2018) develop an ontology to represent different types of e-portfolios and their resources, based on standards and official specifications like IMS ePortfolio and JISC Leap2A. Although these models present a great advance in the use of semantic tools, they do not use the concept of learning paths, a feature that further enriches personalization.
In previous work, authors propose an ontology network called AONet where the portfolio concept has been developed as the container of student assessment and tracking (Romero, Gutiérrez, & Caliusco,
2017,
2018).
The present work extends this network, adding a conceptual model of an intelligent system that uses the concept of learning paths based on portfolios to self-regulate learning. It is based on semantic technologies to provide an automatic generation of learning path taking into account the individual learning paces, capabilities and motivations of students. In this way, the portfolio concept is useful as the tool to collect students’ experiences and their interaction with teacher. Also, the learning level concept allows system to infer recommendation for each student.
The proposed intelligent system will be a useful tool for teachers and students. On the one hand, it will provide recommendations to guide learning activities and select educational resources according the possibilities of each student. This will allow teachers to personalize learning. On the other hand, students will be able to estimate and recognize their learning progress to make decisions that will affect and improve their performance within the possibilities offered by the course.
This work is organized as follow.
Background section presents background about previous works, portfolio, learning path and learning levels.
Intelligence system conceptual model based on semantic technologies section presents the intelligent system conceptual model. It describes Educational Resources ontology and Learning Path ontology proposed. Also, it shows the use of the model through the derivation rules implemented in the ontology and agent rule. Finally, this work is concluded.
Conclusions
In this work, a conceptual model based on semantic technologies of an intelligent system to support self-regulated learning was presented. This model is beneficial for both students and teachers.
On the one hand, the model allows self-regulated learning, enabling students to be proactive in their efforts to learn, know their strengths and limitations, establish personal goals and define strategies to achieve them, monitoring their own progress and improving the methods, they use to learn.
On the other hand, the model allows the personalization of learning to help teachers in the complex task of directing learning process. This personalization consists of the selection of the appropriate educational resources according to the students’ abilities and the objective of the course.
In order to facilitate self-regulation and personalization of learning, the conceptual model introduces the concept of learning path. Each step in the learning path relates the student’s learning level and enables appropriate educational resources to go forward on learning. Here learning level is a data derived from indicators based on the information available in the student’s e-Portfolio. In addition, the conceptual model is based on semantic technologies to provide an automatic generation of learning paths. This model comprises the different learning levels that a student can achieve and metrics that measure progress with the objective that each learner can reach the highest possible learning level.
With the objective of determining the different learning levels, the model incorporates indicators that allow intelligent system to compare among these levels. Agent rules provide recommendations for both student and teacher. These guide learning activities, selection of educational materials according with the possibilities, weaknesses and strengths of each student. In this way, this proposal models the personalization of learning process.
The use of e-portfolios implies carrying out processes of reflection, planning, synthesis, presentation, discussion and feedback, which are part of self-regulated learning. Thus, intelligent system uses all the evidences of progress that are found in the student e-portfolio and recommend when going forward or staying in the same level in the learning path.
The work presented in this paper presents a substantial difference with respect to traditional learning tools developed, since it not only identifies students who do not reach the minimum learning level desired, proposing suitable materials for their improvement, but also students who stand out and show interest, giving them the possibility to face new challenges.
As future work, the authors propose the definition of new indicators that complement the pedagogical perspective of each student in particular and the learning process in general, and also the integration of the ontologies into a software tool for managing the portfolio. In addition, it is proposed the integration of the software with the implementation of the intelligent agents modelled in order to have the recommender system available to assist the learning process in a personalized way and stimulating self-regulation.