Learner Classification Method for Senior Learning with Decision Tree: A Case Study of Thai Senior

Abstract

The Andragogy theory has been suggested that the learning of the senior is different from the child. A senior has more experiences, self-understanding, precise achievement, and advantage in daily life. However, age increasing hurts learning performance. Besides the age factor, other factors can be considered such as gender, education level, internet usage time, health problems, vision problems, hearing loss, memory loss, and congenital disease. Therefore, this study proposes the machine learning-based method for classifying senior learners. In this case, the decision tree model is used. This study is conducted with 60 seniors aged 60–83 years old from the Senior School in Chiang Rai. For this study, four groups of senior learners are determined including Professional, Medium, Less Knowledge, and No Experience learner. The method with scores of performance relied on the learning ability of a person and the method with personal profile factors are studied and compared. The classification results show that the score based and the factor based method provide 95.00% and 91.67% accuracy respectively. Additionally, the results of the factor based method show that the significant factors contributing to learner classification are the amount of the internet using time, memory problems, the number of programs uses, and age respectively.