Abstract:

Students need to acquire not only core knowledge but also the ability to thrive during rapidly changing times. To foster students’ qualities and abilities, teachers need to assess students’ learning by accurately grasping “what kind of qualities and abilities the students have acquired”. To assess them, it is expected to utilize descriptions of students' reflections on learning. We thought that descriptions of reflections on learning will be able to indicate the status of students’ qualities and abilities or the fostering of students’ qualities and abilities.

However, it is not easy to grasp the status of each student’s qualities and abilities from the descriptions in reflections on learning. In addition, it is not easy to select sentences related to fostering qualities and abilities and determine the qualities and abilities the sentence indicates. One promising direction is technologies such as machine learning and text mining, which could be extract sentences related to fostering qualities and abilities from text data of reflections on learning.

The purpose of this study is to develop a system to support grasping the fostering status of their qualities and abilities from text data.

In our approach, we first need to select sentences related to fostering of qualities and abilities from text data (requirement 1). Second, we need to determine the qualities and abilities to which the sentences are applicable (requirement 2). Third, we need to grasp the fostering status of students’ qualities and abilities by using the extracted sentences (requirement 3).

To satisfy the requirements, we designed three functions. Function 1 automatically selects sentences related fostering of qualities and abilities from text data by using a supervised learning model. Function 2 automatically determines the qualities and abilities to which the sentences selected by function 1 are applicable by using morpheme analysis. Function 3 visualizes the fostering status of students’ qualities and abilities on a dashboard by using the sentences extracted by functions 1 and 2.

We developed a system as a Web application with three functions. The flow of use for our system is as follows. First, students describe reflections on learning. Second, our system automatically selects only sentences related to fostering of qualities and abilities by using a supervised learning model. Third, our system automatically extracts keywords related to qualities and abilities by using the user dictionary in morpheme analysis from sentences selected by the supervised learning model and determines the qualities and abilities to which the sentences are applicable. Fourth, our system visualizes students fostering status along with the sentences extracted by functions 1 and 2 on a dashboard. Finally, students grasp the fostering status of their qualities and abilities by looking at the dashboard. And teachers grasp the fostering status of each students’ qualities and abilities by looking at the dashboard.

By using our system, students will be able to grasp the fostering status of their qualities and abilities and adjust their learning through trial and error to decide how to progress. In addition, teachers will be able to assess students learning appropriately and improve their lessons by grasping the fostering status of each student’s qualities and abilities.

In the near future, we will investigate the effectiveness of developed system by using it in actual classes.

Keywords:

Qualities and abilities, descriptions of reflections on learning, supervised learning, text mining, morpheme analysis.