Due to the emergence and flourishing of the Internet, the development of e-Learning systems has become an important research topic in both academia and industries. As a result, many learning systems and learning objects (LOs) were developed. One major problem of these LOs is that they cannot be reused among different learning systems. To resolve the problem, Advanced Distributed Learning Initiative (ADL) developed a reference model called Sharable Content Object Reference Model (SCORM).
There are two kinds of LOs defined in SCORM. One is asset, and the other is sharable content object (SCO). Assets are digital media such as text, images, sound, assessment objects, or any other piece of data. Each SCO is composed of assets or other SCOs. To increase reusability and interoperability of LOs, metadata can be defined for each LO.
ADL developed a metadata model called Content Aggregation Model (CAM). CAM, adapted from IEEE Learning Object Metadata (LOM), classifies all metadata into nine categories, and one of the categories is "RELATION". A relation in the "RELATION" category is mainly used to describe a LO and express relationships between learning resources. When used skillfully, a relation is a very useful metadata that can enhance learning effectiveness as well as increase the reusability of LOs. For example, as shown in Figure 1, [LO.sub.A] describes how bubble sort works. At the bottom of [LO.sub.A], there is a figure illustrating how bubble sort works in steps. Using the proposed relation metadata, we can define the figure as a learning object of type "Illustration". When the figure is stored in a repository, it can be easily searched and reused by other users. Additionally, the application of relations can be further extended. If the author of [LO.sub.A] wishes to provide more illustrations to help learners, she can easily provide links to more illustrations such as [LO.sub.B] and [LO.sub.C]. LOs such as [LO.sub.B] and [LO.sub.C] can be created by the author or other authors as long as they can be accessed. Also, these LOs can be searched and retrieved from repositories.
As defined in CAM, there are twelve suggested values as shown in Table 1 for the "RELATION" category. However, these suggested values can only describe structure-oriented relationships and cannot express semantic relationships between learning resources (Steinacker et al., 1999). In the past, many relations were proposed (Karger et al., 2006; Ullrich, 2004, 2003, 2005; Loser et al., 2002; Steinacker et al., 1999, 2001; Sddik et al., 2001; Fischer, 2001; Fischer et al., 1999). These relations are mainly based on two major theories. One is instructional design theory (IDT), and the other is rhetorical structure theory (RST). Although these relations can express meaningful relationships between LOs, they are limited in the following ways.
[FIGURE 1 OMITTED]
First of all, some of the relations are redundant. For example, both "Example" and "Illustration" relations are not only defined in IDT, but also defined in RST. Secondly, several relations have already been defined in CAM. For example, the "Interactivity" relation defined in IDT has been defined in the "EDUCATION" category of CAM. Thirdly, although the relations can express semantic relationships between LOs, the usefulness of the relations has never been formally studied. Fourthly, there is no learning content management system supporting new relations. Finally, after studying the relations, we found that some relations are confusing and may not be easy for authors to clearly specify relation for each LO.
In this paper, we studied all 32 relations proposed in IDT and RST, removed 8 duplicated relations, and analyzed the rest 24 relations. A web-based learning content management system supporting these 24 relations was designed and implemented. By using the system, authors are allowed to assign a specific relation to a LO. …