Building knowledge structures for online instructional/learning systems via knowledge elements interrelations

Abstract

The knowledge structure of an online instructional/learning system can serve as a guide that directs an inexperienced learner to appropriately capture knowledge. In this paper, a Knowledge Construction Model (KCM) that helps instructors in any domain to build a knowledge structure for an online instructional/learning system is designed and developed. The developed KCM contains two parts. The first part is to preprocess the knowledge elements and their interrelations for a defined subject provided by an instructor or instructors. Based on the interrelations of knowledge elements, the second part is to generate the knowledge structure on a level-by-level basis. A practical application case that is used to demonstrate the developed KCM is delineated.

Introduction

The OnLine Instructional/Learning Systems (OLI/LS) have been successfully used to provide the learners with an environment of one-to-many instruction, learning without limited time and place, different kinds of knowledge sources, open learning systems, information delivery with multimedia format, cooperative learning, and storing and quickly accessing huge amount of learning materials (Burbules and Callister, 1996, Penfield and Larson, 1996, Sun and Chou, 1996, Paolucci, 1999, Grabe and Sigler, 2002). In order to have the online materials to be reusable, accessible, durable, interoperable, adaptable, and affordable, the organization of Advanced Distributed Learning Initiative (ADL) has designed and developed a model (Sharable Content Object Reference Model, SCORM) to help produce OLI/LS materials (http://www.adlnet.org). Basically, the SCORM contains two major parts. One is content aggregation model (CAM) and the other is run-time environment (RTE). In CAM, particularly, the strategy that SCORM takes on for managing learning materials is to define, design, and develop standardized learning objects as knowledge elements of learning subjects. Thereafter, when producing an OLI/LS course, an instructor can pick up some (or all) of these learning objects and then organize them into a course package. Importantly, it has been seen that the SCORM do provide the OLI/LS with a meaningful methodology in modeling learning materials (Huang et al., 2002).

However, it is generally believed that to arrange knowledge elements with respect to the sequences for an OLI/LS is a complex, but important task to learners, in particular to a beginner. When using an OLI/LS, a beginner may not have the basic knowledge about what is the next step he/she should go, and thereafter may travel several miles in width, but 1 in. in depth. Consequently, major issues that include low learning effectiveness, loss of learning objective, and loss of learning direction may weaken the capability of an OLI/LS. A study conducted by Paolucci (1999) indicated that a relation do exist between the structuring of the knowledge domain and positive learning performance. More recently, Hoogveld, Paas, Jochems, & van Merriënboer (2001) pointed out that instructional design systems-based training produced a better training result in comparison with the experienced-based design condition in their empirical investigation. Consequently, a structure with respect to the arrangement of knowledge elements that deals with the instructional design problem becomes necessary in the OLI/LS development life cycle, and thus is the motivation of this research.

Eventually, the goal of an OLI/LS is to help learners obtain the whole domain knowledge rather than pieces of knowledge. In order to improve the capability of an OLI/LS, researchers pay increased attention to instructor modeling that deals with defining knowledge elements, ordering knowledge elements, grouping knowledge elements, and appropriately directing learning paths (Jonassen and Hannum, 1995, Hackbarth, 1996, Newby et al., 1996, Ferry et al., 1998). Therefore, a structure for knowledge elements becomes essential due to the requirements of organizing the massive amount of learning materials. Gordon (2000) suggested that creating a knowledge structure needs to identify specific pieces of knowledge and consider their relationships. Major techniques that can be used in the process of instructional design include selecting and sequencing (Landa, 1983, Gagne, 1985, Scandura, 1987, Jones et al., 1992, Hackbarth, 1996, Mengel and Adams, 1996). Particularly, the sequence of knowledge elements for a subject is often implied by the nature of knowledge in the element description. However, it is more than simply a description of the sequence in which the knowledge is performed (Jonassen & Hannum, 1995). The sequence for performing the knowledge implies an appropriate instructional progression. Consequently, the representations of materials have to be leveled by defining the interrelations among the connections. For example, knowledge of relational integrity rules in the subject of database management systems has to be covered before normalization is given. Therefore, the problem of ordering the conceptual links in sequence becomes critical while building the knowledge structure for an OLI/LS.

The concepts of breadth-first algorithm and depth-first algorithm that are widely used in the application of the artificial intelligence realm are utilized to organize the links of state spaces (Luger & Styblefield, 1993). The breadth-first algorithm uses a strategy that explores the space in the level-by-level fashion while depth-first goes deeper into the search space whenever it is possible. These algorithms are eventually used to determine the order where states are examined in a tree or graph. Therefore, the knowledge structure is basically formed before these algorithms can be performed. On the contrary, instructional design deals mainly with the formation of a knowledge structure for the knowledge elements defined. The Depth-First Structures (DFS_TRUCTURE) and Breadth-First Structures (BFS_TRUCTURE) that are used to help construct knowledge elements are employed. Any two of the knowledge elements with relations of dependency can form a DFS_TRUCTURE. A BFS_TRUCTURE forms if no relation for two elements is defined. The entire knowledge structure is built via the integration of BFS_TRUCTURE and DFS_TRUCTURE. The study presented in this paper is to develop a Knowledge Construction Model (KCM) that can help efficiently build a knowledge structure that guides a beginner to appropriately capture domain knowledge on a level-by-level basis.

The remainder of this paper is organized as follows. The DFS_TRUCTURE and BFS_TRUCTURE are described in Section 2. Section 3 presents the developed KCM where the user interface, knowledge table, model mechanism, and update facility are included. A practical application case used to demonstrate the utilization of KCM for the subject of database management systems is delineated in Section 4. Section 5 addresses the research conclusion.