The Viability of SSM as a Novel Technique for Knowledge Elicitation and Structuring in a Complex and Unstructured Virtual Learning Environment (Research Paper Sample)

The Viability of SSM as a Novel Technique for Knowledge Elicitation and Structuring in a Complex and Unstructured Virtual Learning Environment
Franklyn CHUKWUNONSO1
fchukwunonso2@live.utm.myRoliana IBRAHIM2
roliana@utm.myAli SELAMAT3
aselamat@utm.my
Author(s) Contact Details:
1,2,3 Department of Information Systems, Faculty of Computing, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia

Abstract— The advent of advanced Internet and web-based technologies have significantly increased the important role played by technology in e-learning such as faster and effective transmission of knowledge, but has also at the same time compounded the complexities associated with the use of technology enhanced learning (TEL), such as unstructured and excessive information content. This calls for the need to redefine the role of knowledge in e-learning; how it can be elicited, represented, and delivered to learners within their context of use. This study attempts to address these challenges by proposing a novel approach for knowledge elicitation and representation through the use of Soft Systems Methodology (SSM) for eliciting and structuring knowledge in a very dynamic and unstructured virtual learning environment (VLE). The results indicate that SSM is a viable and successful tool useful for rich knowledge elicitation and problem structuring and is best suited for complex, unstructured, dynamic, and poorly defined problem cases such as in VLEs. The major contribution of this study is that SSM can be employed for eliciting and structuring knowledge for the development of a knowledge-based personalized e-learning system.
Keywords – Knowledge-based Systems; Personalized E-learning; Soft Systems Methodology; Virtual Learning Environments
1. INTRODUCTION
The sharing of knowledge whether in the office, classroom, street or media has been an existing phenomenon since the creation of man. Nevertheless, there is a major distinction between knowledge-giving (as in the aforementioned) and knowledge-sharing. Knowledge-sharing is a complex task that involves the exchange of knowledge in a two-directional way and is often characterized by practical barriers that needs to be identified and overcome. Some of these barriers become more pronounced when the parties involved in the exchange of knowledge are confined in a spatially distributed domain-specific context. Some studies have identified these barriers to include perceived power loss (Kuo et al., 2014), job insecurity (Chae et al., 2014), lack of motivation (Nissen et al., 2014), resistance to change (Durmusoglu et al., 2014), and inconsiderate language use (Park and Lee, 2014). Most of these barriers are soft issues common with any human activity system. These soft issues are always an obstacle to knowledge elicitation. Furthermore, knowledge models help teachers’ structure different knowledge types based on their relationships within the e-learning context in order to achieve personalization of learning resources. This requires the teacher to explicitly understand the different roles knowledge plays in constituting a learning content based on the individual characteristics of each knowledge fragment (Koh and Chai, 2014; Chukwunonso et al., 2013). This follows the argument of Moustaghfir and Schiuma (2013) that tactic knowledge, unlike explicit knowledge, cannot be easily identified, quantified or communicated as it is not explicit or formal but simply an individual’s conceptual and cognitive understanding of external processes. Explicit knowledge however relies on objectivity, is convenient to express and can be communicated because of its formal structures that have been proven over time (Biggam, 2002; Bakardjieva and Galya, 2011). Thus, the ability to identify what knowledge to elicit and properly highlighting the type of barriers that pose a challenge to eliciting that knowledge, remains a major challenge to current knowledge elicitation methodologies (Chukwunonso, et al., 2014). The description used by Newell (1982) at the knowledge level has since been in use as the base principle on which knowledge elicitation is founded, and has provoked several studies in this regard. The need to the shift away from rigid agile methodologies to softer and more appropriate methods that can take into consideration the complex and unstructured social contexts under which these virtual learnings take place, has been long overdue (Chukwunonso et al., 2014; Motta, 2013; Howell et al., 2010). This paper therefore proposes SSM as a viable methodology for knowledge elicitation appropriate for constructing KBSs in dynamic and uncertain virtual learning environments, given its very flexible methodology, which can be adapted to suit the nature and context of the problem being solved. The rest of the paper is structured as follows: Section two takes a look at the background of the study, Section three presents the methodology adopted for this research, Section four discusses the results and presents the proposed knowledge-based personalized e-learning model and Section five concludes the paper, stating its limitations and areas of future study.
2. STUDY BACKGROUND
Knowledge elicitation is a very important aspect of knowledge-based systems (KBSs) development. Knowledge elicitation is the process of extracting domain specific knowledge with regards cognitive issues underlying human performance (Diaper, 1989; Wang, 2013). First used in expert systems and now in KBSs, several studies have been seeking the best approach to developing knowledge bases into applications or systems such as intelligent tutoring systems, training or educational systems, expert systems, adaptive user-interfaces, e-learning, etc (Chukwunonso et al., 2013). In other words, knowledge elicitation is a sub process of knowledge acquisition, even though the two are interchangeably used, and all of which make up KBSs. The new drive for KBSs have raised concerns in both basic and applied sciences on the best way in which knowledge can be effectively and efficiently elicited from an expert (McAndrew and Gore, 2013; Chiravuri et al., 2011). Early studies on knowledge elicitation focused on direct extraction but this was quickly put aside because of the complexity that characterized the problem context (Jetter and Kok, 2014). Other limitations of early knowledge elicitation techniques include bias, error and flawed verbal reporting of experts (Clark et al., 2012). McAndrew and Gore (2013) criticized the use of cognitive theory for knowledge elicitation stating that although it addresses the knowledge issue, the aspect of its representation and varied conceptualization of the structure of the knowledge (schemata, prototypes, semantic networks, etc.) were not addressed, and which of course, is the main focus of KBSs. Recent approaches to knowledge elicitation make use of model constructs to reflect the expert’s knowledge (Starr and de-Oliveira, 2013), with focus on formal and symbolic representation of knowledge, and how such representations are actually obtained.
During the last three decades, several studies have been devoted to seeking better and more efficient ways of developing KBSs. From the early days of Artificial Intelligence (AI) and Expert Systems (ES), to the more recent Knowledge Engineering (KE), several methodological approaches have been developed and tested on how best to elicit and transfer expert knowledge for the construction of KBSs. With the proliferation of information due to the daily advances being recorded in Internet and web technologies, the traditional concept of “knowledge role” metamorphosed from static (symbolic level) to a more dynamic (knowledge level), thereby requiring new methods of knowledge elicitation and representation techniques (Chukwunonso et al., 2013). The introduction of ontology attempted to address this challenge at the knowledge level but was inadequate and inefficient for the construction of KBSs in a complex and unstructured networked environment (Motta, 2013). Although still evolving, SSM can arguably be said to be the most appropriate method for constructing KBSs through a KE domain-specific problem-solving approach (Zaied, Abdel and Hassan, 2013). SSM has been described in several ways by different studies. First proposed by Checkland (1981), and Checkland and Scholes (1990) as an alternative method for solving soft information systems development issues. SSM addresses both structured and unstructured problem cases. Although generally regarded as a learning system, its wide applicability and acceptability has largely been credited to its ability to address soft unstructured problem situations in environments characterized by high levels of uncertainty and complexities, which is a common phenomenon in most human activity systems. SSM makes use of knowledge concepts to highlight non-technical and fuzzy issues in web environments (Herman, 2011). This approach is based on the theory of systems thinking where the analyst or knowledge engineer leaves behind his preconceptions about the problem situation and tries to define the problem based on a world view (weltanschauung). The theory assumes that everybody is right based on a theoretical perception of the problem and attempts to unify all the divergent views into a Weltanschauung where all the actors share a common view through a seven stage process (see Figure 1). The problem Checkland tried to address was how to represent expert ideas in an unstructured problem situation into a structured one by making use of rich picture constructs, Root Definitions, CATWOE and models to represent the real world problem so as to elicit the expert’s knowledge. SSM is based on the argument that every individual in a human activity problem situation have different perceptions of the problem, approaches to addressing it, benefits and risks derived as a res...