Analysis of the influence of smartphone overdependence using machine learning focus on imbalanced data†
Kwang Yoon Song1 · Youn Su Kim2 · In Hong Chang3
13Department of Computer Science and Statistics, Chosun University
2Department of Computer Science and Statistics, Graduate School, Chosun University
2Department of Computer Science and Statistics, Graduate School, Chosun University
Correspondence to: † This study was supported by research fund from Chosun University, 2022.
1 Assistant professor, Department of Computer Science and Statistics, Chosun University, Gwangju 61452, Korea.
2 Ph.D. candidate, Department of Computer Science and Statistics, Graduate School, Chosun University, Gwangju 61452, Korea.
3 Corresponding author: Professor, Department of Computer Science and Statistics, Chosun University, Gwangju 61452, Korea. E-mail: ihchang@chosun.ac.kr
1 Assistant professor, Department of Computer Science and Statistics, Chosun University, Gwangju 61452, Korea.
2 Ph.D. candidate, Department of Computer Science and Statistics, Graduate School, Chosun University, Gwangju 61452, Korea.
3 Corresponding author: Professor, Department of Computer Science and Statistics, Chosun University, Gwangju 61452, Korea. E-mail: ihchang@chosun.ac.kr
Received July 19, 2023; Revised August 30, 2023; Accepted September 18, 2023.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
- Abstract
- Smartphones provide us with convenience, but they also make us very dependent on them. As much as we rely on them, excessive smartphone use can cause physical and psychological problems in humans. This is called smartphone overdependence, and there is a lot of research on it. In this study, we proposed a smartphone dependency classification model using random forest, a machine learning technique, for the entire survey population based on survey data from the smartphone dependency survey conducted by the National Information Society Agency. Since the data on smartphone overdependence is imbalanced and the percentage of smartphone overdependence is very small compared to the general users, we utilized the oversampling technique to overcome this, and showed excellent classification rate, sensitivity, and specificity by increasing the percentage of smartphone overdependence by 2, 3, and 4 times. We found that the most influential variable in the classification of smartphone dependence was the use of smartphones for leisure pursuits such as playing games, watching movies, and watching videos. We may all be heading towards a naturalized smartphone dependency. We may need to relax the strict criteria for smartphone dependence in the past to accept and improve the current situation.
- Keywords : Imbalanced data, oversampling, random forest, smartphone overdependence
Full Text(PDF) Free
