Nabila Bousbia

Which Contribution Does EDM Provide to Computer-Based Learning Environments?

Educational Data Mining is a new growing research area that can be defined as the application of data mining techniques on raw data from educational systems in order to respond to the educational questions and problems, and also to discover the information hidden after this data. Over the last few years, the popularity of this field enhanced a large number of research studies that is difficult to surround and to identify the contribution of data mining techniques in educational systems. In fact, exploit and understand the raw data collected from educational systems can be “a gold mine” to help the designers and the users of these systems improving their performance and extracting useful information on the behaviors of students in the learning process. The use of data mining techniques in e-learning systems could be very interesting to resolve learning problems. Researchers’ ambition is to respond to questions like: What can predict learners’ success? Which scenario sequence is more efficient for a specific student? What are the student actions that indicate the learning progress? What are the characteristics of a learning environment allowing a better learning? etc. The current feedback allows detecting the usefulness of applying EDM on visualizing and describing the learning raw data. The predictions take also an interest, particularly the prediction of performance and learners’ behaviors. The aim of this chapter is to establish a bibliographic review of the various studies made in the field of educational data mining (EDM) to identify the different aspects studied: the analyzed data, the objectives of these studies, the used techniques and the contribution of the application of these techniques in the field of computer based learning. The goal is not only to list the existing work but also to facilitate the use and the understanding of data mining techniques to help the educational field specialists to give their feedback and to identify promoter research areas in this field to be exploited in the future.

Indicators for Deducting the Learners' Learning Styles: Case of the Navigation Typology Indicator

Research in individual differences and in particular, learning and cognitive style, has become a basis to consider learner preferences in a web-based educational context. How learner’s learning style influences his/her navigation behavior has been investigated by several studies, which indicate that we can deduce the learning style from the navigation behavior. In this paper, we propose an indicator of “navigation typology”. We detail the way in which this indicator is calculated, based on tracks analysis, which are aggregated into low and intermediate level indicators to determine the value of the navigation typology.

Detection of Learning Styles from Learner's Browsing Behavior During E-Learning Activities

One of the bases of adaptation and learning tracking is the learners modeling. Research in this field, or more generally in the field of user modeling, was sustained mainly on the detection of features related to the users knowledge, interests, goals, background, and individual traits [3]. We are interested in this last aspect, in particular the identification of the learning style. In this paper, we propose an approach for the learners activity perception on an e-learning platform to identify the users learning styles from observable indicators related to their learning path and interactions.

Adaptation Based on Navigation Type and Learning Style

In this paper, we present an adaptation approach of e-learning content based on the navigation type indicator describing the learner’s behavior while browsing an e-learning course. This adaptation approach benefits from the found correlation between this indicator and learning styles, particularly Sequential/Global and Active/Reflective styles. Many studies use leaning styles for adaptation based on educational rules. Thus, we propose for each value of the navigation type indicator, to provide the learner with the appropriate adaptation to the learning style correlated with the indicator value.

A User-Centered Educational Modeling Language Improving the Controllability of Learning Design Quality

Semiformal EMLs are developed to facilitate the adoption of educational modeling languages (EMLs) and to address practitioners’ learning design concerns, such as reusability and readability. In this article, SDLD (Structure Dialogue Learning Design) is presented, which is a semiformal EML that aims to improve controllability of learning design quality, one of the major concerns of practitioners that has not been previously addressed by EMLs. To address this problem, SDLD follows the proposals of Transactional Distance Theory (TDT). The adoption of this theory helps to measure, identify, and integrate the determinants of learning design quality in the creation of learning designs. Results from using SDLD showed that it improves the expressiveness of EMLs, especially in terms of the interactional facet of learning units and confirmed that it improves the usefulness of existing EMLs. In this context, SDLD represents, on the one hand, a first step toward the operationalization of the TDT. On the other hand, it is another way to facilitate the adoption of EMLs for their utility.

A Semantic Analysis of the Learner’s Disorientation

Learning systems are dedicated for learning about a particular area organized through non-linear documents. Therefore, it is always useful to recognize the state of knowledge and the navigation behaviour of a learner in order to evaluate customize and adapt the learning process. In this paper, we aim to make a semantic analysis of the learner’s navigation during his apprenticeship in hypermedia content. The main reason of this analysis is to identify the browsing behaviour of a learner with the current course. We assume that if the semantic distance between the domain concepts of the documents (or pages) that follow each other in the navigation of a learner is great, then this reflects the unstructured navigation behaviour and interprets that the learner is disoriented. This type of behaviour could be due to the poor organization of the content and the bad structuring of the course. Indeed, this analysis will allow to the tutor, to identify the disoriented learners and help them, and to the course author, to specify the causes to restructure and to deepen the analysis of the existing content and the navigation links between the parts of the course.

Modeling approach of generic courses

We present in this paper a modeling approach of the pedagogical sequencing of learning units which is called a PGM Pedagogical Graph Model. This graph allows expressing the totality of the pedagogical constraints. PGM is a graph in which the nodes are the learning units and the arcs are the pedagogical constraints between learning units. This will help the author to well structure his/her course and pedagogical relations between the various pedagogical units that compose it and help the learner to be oriented during the browsing in the course avoiding thus the disadvantages of the hypertext links.