Abstract:

In recent years, personalized learning has been receiving attention for fostering student skills and competences. Personalized learning is teaching and learning that focuses on the learner’s background, needs, potential, and perceptions. This is known as learner-centric education. To achieve personalized learning, it is important for students to continually assess their own learning situations from various perspectives, such as their own growth and learning outcomes.

On the other hand, learning analytics (LA) has also been attracting attention in recent years. LA refers to analytical and data mining methods specialized for use in educational datasets. Using LA enables students to grasp their own learning situations, and it is expected that students can do so from a variety of perspectives by using various LA.

However, little research has been done on the development of a system that can switch between various LA.

The purpose of this study is to develop a system that enables students to switch between various LA while continuing to learn. In this study, we proposed a model for designing a system with a module that can switch between various LA. Next, we developed a system that was designed on the basis of the model.

To develop the system, first, it is necessary to be able to analyze and visualize educational data by using various LA (Requirement1). Second, it is necessary to be able to switch between various LA while the system is running (Requirement2).

We proposed a model for designing a system that satisfies these requirements. A system designed on the basis of the model mainly consists of four components.
Database : This enables the system to register or manage educational data.
Interface : This enables users to use the system.
Module to conduct various LA (corresponds to Requirement1) : This enables the system to execute processes for various LA. Specifically, this can register programs for analyzing or visualizing educational data for various LA. In addition, this enables those programs to analyze educational data in the database and visualize the results on the interface.
Module to switch various LA (corresponds to Requirement2) : This enables the system to switch between various LA. Specifically, when a user requests any LA via the interface, this calls a program corresponding to the LA from the target module to conduct various LA.

We developed a system designed on the basis of the model as a Web application. The system has three functions.
Function1 : This function uses supervised learning to predict a student’s learning situation about competencies to learn tenaciously and autonomously.
Function2 : This function uses unsupervised learning to group all students in a learning community into five groups in similar learning situations and visualizes the analysis results as a scatter plot.
Function3 : This function visualizes the analysis results of Function1 and Function2. In addition, this function enables students to switch between those.

We conducted an operation verification test of our system to verify that it works properly. In this verification, we first created dummy data and imported it into system’s database. We then verified that all functions of our system worked properly using the data.
The results of the verification confirmed that all functions of our system work properly.

In the near future, we will evaluate our system and investigate its effectiveness in detail.

Keywords:

Learning Analytics, Machine Learning, Educational Data, Module to Switch between Various Leaning Analytics, Personalized Learning.