Proposal

Introduction


Overview


Online learning is viewed as an effective way for distributing knowledge ubiquitously, especially with the robust development of Massive Open Online Courses (MOOCs) by academic institutions such as Stanford University's edX and Coursera. From enrollment to managing class discussions and assignments, these MOOCs succeed in migrating the classroom entirely to the cloud. The same concept has been applied to distant or blended learning, where learners are able to interact with materials uploaded online. Through such interactions, learners undergo a process of self motivated learning.

Self motivated learning (SML) describes the way in which an eager learner incorporates various self-regulation processes with task strategies and self-motivational beliefs. These learners go through their learning processes in three cyclical phases, namely: forethought, performance control and self-reflection Zimmerman 2000. In the second phase, learners gather information to evaluate the effectiveness of the strategic plan imposed and improve future attempts of learning. This phase is crucial as research shows students who are trained in self regulation processes display high levels of achievement Schunk 1996. In light of this view, some educators pointed to knowledge graphs for facilitating sharing of learning content.

In experiments led by researchers, learners are encouraged to use cognitive structures (i.e. knowledge graphs) to help themselves interpret, organize and share their knowledge. The experimental results show increased qualitative performance as well as increased engagement among both learners and teachers Chu 2010 Hwang 2011.

While current learning management systems lack promotion on SML, learners tend to spend unnecessary amounts of time on skimming and searching for designated information. The most obvious reason is that those learning management systems serve merely as repositories of information, rather than a catalyst for fostering SML activity. Therefore, a comprehensive learning management system that has an interface to create knowledge graphs for course material is needed for motivating self motivated learners, and ultimately for improving online learning throughput.

This paper aims at two contributions. First, it provides an implementation of the knowledge graph cognitive structure that integrates as a learning management system. Second, it proposes several interactive plug-ins for the creation and management of learning assets and evaluation of learners' performance.

Objectives

The main goal for this paper is to implement a novel learning management system that incorporates knowledge graphs for the creation of learning assets that can be categorized and accessed by topic. These learning assets can be annotated with metadata and be linked-up with plug-ins for more interactive use.

The following goals will be focused on during project development:

1. To provide an interface to access, modify and delete knowledge graphs for different user groups
2. To build a file manager that links file snippets to knowledge graph nodes
3. To build an interactive Question Box for learners to ask questions and teachers to answer them
4. To develop a Quiz System that attaches short quizzes to certain learning materials
5. To provide analytics on student performance and behaviour for administrators or teachers

To achieve goals #1 and #2, specific data structures and data retrieval techniques will be utilized.

To achieve goals #3 and #4, user interfaces that motivates participation will be studies and incorporated.

To achieve goal #5, statistical methods will be applied to show relationship between data sets.

The biggest challenge is to maintain a similar level of interactivity across different screen sizes. For example, the teacher can drag nodes on a knowledge graph to connect them on a touchscreen.


Literature Survey


A range of learning management systems (LMS) , both commercial and open-source, are available to educators. This section will discuss the pros and cons of a few popular LMS, including LMES, the system previously used by HKUST, Canvas, a commercial learning management solution currently used by different institutions such as HKUST Instructure, Moodle, an open-source LMS used by nearly one million users Moodle in the world, in terms of the way how the course materials are managed and presented, and the important plug-ins that facilitate effective learning.

Repository of Materials

One of the feature most commonly used by student is the repository of course materials Costa 2012. In LMES and Canvas, this feature is implemented as file systems where the course instructors can upload files such as presentation slides and supplementary notes to the course directory accessible to but not modifiable by students. While file systems are easy to implement, from the learner’s perspective, it may take some time to find materials that are unorganised, as Figure 1 shows. Furthermore, due to the tree-like structure of file systems, it is impossible to link up the related content except by putting them under the same directory. As a consequence, students may need to search the whole directory just to locate, for example, the prerequisite information for a particular topic.