PewePro 2

Domain Model

Domain model of an adaptive educational course defines domain conceptualization (knowledge to be learnt) along with its connection to educational content. Domain model in ALEF is represented by two distinct layers – metadata layer and educational content layer – connected via various types of relationships.

A schema of the domain model of ALEF.

A schema of the domain model of ALEF.

Educational Content Layer

The term content refers mainly to a learning object as to fundamental course content representation. However, in a collaborative and read-write environment we support also other types of content entities such as blog (created by both students and teachers) or discussion thread (composed of interlinked comments added by students).

Learning objects are represented by entity LearningObject and are further divided into the following types:

Different learning object types and their seamless combination is an important distinguishing feature of our approach. The student is not required to read explanations in one system and practice the acquired knowledge in another one. Different forms of interaction within one system improve student's learning experience and learning outcome.

Explanation represents instructional content that describes a subject domain. Question and Exercise represent interactive part of a course. Questions aim to provide students with an immediate feedback on their knowledge with further explanation of respective answer choices. System supports various types of questions:

Other question types can be easily added by subtyping the Question entity. Exercises allow students to practice gained knowledge. Besides enriching the learning process, gathered information is also used to update student model to keep track of user’s knowledge.

The domain model covers two distinct parts of every learning object:

The chosen representation allows for an easy definition of other types of learning objects. The actual content goes to the textual files while its metadata comes to the model as a new subtype of LearningObject.

Learning objects are composed of reusable parts referred to as fragments. Fragments are smallest pedagogically coherent units allowing content-based adaptation. Fragments are further subtyped according to a type of particular learning object:

Explanation can be composed of one or many Definition fragments. Exercise can be composed of Definition, Hint, Solution and/or Clarification fragments. Similarly, question can be composed of Description and Answer fragments. Each fragment type has its pedagogical role in learning flow and “assists” a student to achieve his or her learning goals.

The order of fragments and metadata related to fragment usage within a learning object is defined by an entity FragmentUsage.

Metadata Layer

Every content entity is associated with respective metadata. Comparing to other existing approaches, the notion of metadata in ALEF is quite simplified in order to achieve the degree of complexity, which is manageable by ordinary users (both teachers and students). This allows for automatic construction of domain model, and, on the other hand, it provides a solid basis for reasoning resulting in advanced operations such as metadata-based navigation recommendation.

We consider following metadata types:

The basic metadata entity is Concept. It is a domain knowledge element usually representing topic or subtopic contained within the content (learning object). Concepts are mutually interconnected and form a structure, which typically resembles ontology. Different types of relationships have different semantics: e.g., concept relatedness, hyper-/hyponymy, prerequisite. Concepts are also associated with course content represented by learning objects.

Annotation represents additional content of a learning object or its fragment, which can be created by students or teachers. It is subtyped to support many different types of annotations:

Metadata can be interconnected by different types of weighted relationships. We defined ConceptToConceptRelation, which represents fundamental connection between any metadata entities. Other types of relations (e.g. prerequisite) are defined by extending this relationship type.

In ALEF, both content entities and metadata entities are considered resources. This generalization is based on the fact that the boundary between content and metadata is very thin (e.g. a comment assigned to learning object can be itself assigned other comments resulting into discussion thread formation). Second reason to generalize Content and Metadata into Resource is the way we model interaction in the framework – relationships between user and content and between user and metadata are very similar. Figure shows diagram representing described ALEF domain model.

ALEF domain model.

ALEF domain model.