EDUCATIONAL DELIVERY
7.4 What is Blended and/or E-learning?
7.4.2 Adaptive E-learning
The need for personalized access to information has become a well established feature of many application areas in ICT (Brusilovsky & Henze, 2007). Areas such as e-commerce, news access, and many others already successfully make use of a variety of adaptive features. However, despite the fact that educa-tion, as a field, has one of the biggest needs for personalized access, adaptive technologies have not yet become widely used in education (Brusilovsky &
Henze, 2007). This is primarily due to the fact that most of the ”adapta-tion techniques that focus on user interests and work successfully in other fields have a limited applicability in the educational context” (Brusilovsky
& Henze, 2007). In education users do not only differ in terms of interests, but also requires adaptation based on current knowledge levels, goals, skills, and personal learning styles (Brusilovsky & Henze, 2007). These adaptations form the basis of adaptive e-learning technologies.
Adaptive e-learning systems, as well as intelligent tutoring systems, are categories of e-learning solutions that make use of advances from the field of artificial intelligence (AI) to better support learning. The primary purpose of adaptive e-learning systems is to automatically adapt presented learning content to the learner. Artificial intelligence can be seen as the key to this
automation of e-learning (Hentea, Shea, & Pennington, 2003). However, for the purposes of this thesis, the exact ”form(s)” of artificial intelligence used need not be examined. Instead, the underlying architectural structure that enables the adaptation of content in current adaptive systems will be briefly examined.
Current adaptive e-learning systems make use of a user model and a do-main model to provide adaptive features. ”The user model is a representation of information about an individual user that is essential for an adaptive sys-tem to provide the adaptation effect, i.e., to behave differently for different users” (Brusilovsky & Mill´an, 2007). The adaptive system collects knowl-edge about the user from various sources. These sources could include both implicit observation of the user’s behavior and/or explicitly asking the user for inputs (Brusilovsky & Mill´an, 2007). The type of information represented in a system’s user model, as well as how much information is stored, will de-pend on the specific kind of adaptation effect that the system has to deliver (Brusilovsky & Mill´an, 2007). Adaptation of educational resources is done in order to increase the usefulness of the educational resource to the learner.
How useful a specific resource is to a learner depends on many factors.
For example, ”some resources may require additional knowledge that the learner does not yet have (in accordance to his/her user model), while others may teach the subject without sufficient in-depth information and are thus too easy for this learner” (Brusilovsky & Henze, 2007). The learner’s current progress, personal learning style preferences, learning goals, and many other factors could thus all form part of the user model in a specific adaptive e-learning system. It should be clear that the user model will be continuously updated and adapted as the learner progresses through an e-learning course.
The user model is used in combination with an expert-, or domain model.
The domain model is also sometimes referred to as the ”ideal student model”
(Brusilovsky & Mill´an, 2007). This domain model is constructed by subject experts and reflects the knowledge an ”ideal” learner should have after suc-cessfully completing the desired educational module. The domain model will also contain a large amount of meta-knowledge regarding the pre-requisite knowledge that a learner would need before being able to attempt a spe-cific learning task, and other relationships between various units of knowl-edge. Before choosing specific content to present to a learner, an adaptive
system performs a comparison, with the help of artificial intelligence based techniques, between the user model and the domain model. Some systems perform this comparison with the assistance of an overlay model. An overlay model represents an individual user’s knowledge as a subset of the domain model. ”For each fragment of domain knowledge, an overlay model stores some estimation of the users knowledge level of this fragment” (Brusilovsky
& Mill´an, 2007). Through the use of these user-, domain, and sometimes overlay-, models, an adaptive system can determine how to adapt the actual content of a learning module to be more applicable for a specific learner.
However, the use of these models is not restricted to only the adaptation based on current knowledge.
User models can also be used to contain information specifically focused on user interest. In many current adaptive systems modeling user interest is seen as more important than modeling user knowledge (Brusilovsky & Mill´an, 2007). In an information security context, this might not necessarily be the case. Even if a user is not really interested in learning about information security it would still be necessary for the organization to ensure the user is educated. However, the same techniques that are used to model interest could probably be applied to model the applicability of specific knowledge based on a user’s security roles and responsibilities. Brusilovsky and Mill´an (2007) also discusses the use of user models to model goals and tasks. This form of adaptation depends on the current context within which the user operates and would be used to present information that could fill the users immediate information needs. As an example, due to such a contextual adap-tation, a user actively engaged in a task where encryption techniques could be of importance, might be presented with knowledge specifically relating to the use of encryption techniques. Additionally, user models can be used to model user background (previous experience outside the core knowledge domain), individual traits (including cognitive styles and learning styles), and work context (location, systems/platform(s) used, etc) (Brusilovsky &
Mill´an, 2007). Adaptation of e-learning materials based on such extended user models would obviously require corresponding changes in the supporting domain models. As a field of study, adaptive e-learning is constantly evolving and improving.
According to Brusilovsky and Henze (2007) current adaptive techniques
can already be used
• ”to support the learner in finding the most appropriate learning re-source;
• for providing awareness about the learning process (e.g., by pointing out necessary pre-knowledge that this learner might otherwise miss);
• for providing guidance (e.g., by providing an individually tailored se-quence of learning resources which teach the topics s/he is interested in while incorporating all required prerequisite knowledge);
• for providing orientation (e.g., by pointing out the next learning steps to take, or the existence of different schools-of-thought);
• for considering individual learning styles” (Brusilovsky & Henze, 2007) Adaptive technologies already play an important role in e-learning and will probably start to play a progressively more important role in future e-learning solutions. It is the opinion of the author of this thesis that these technologies can play an especially important role in the use of e-learning for organizational information security education. This possible use will be discussed later in this chapter.
This section established what e-learning and blended learning are. It also explored the components that commonly form part of such systems, and the role(s) these components play in such a system. Finally the field of adaptive e-learning was briefly introduced and discussed. The next section will specifically focus on the benefits e-learning, and/or blended learning can offer to organizations.