Journal of Advanced Mechanical Design, Systems, and Manufacturing (Mar 2021)
Proposal of fingering detection method for touch typing learning support system
Abstract
Recently, typing skills especially touch typing is given much importance. However, the typing skill of entrant students tends to be poor, because of the use of smart phone. Although some students try to improve their typing skill using typing software, pre-existing one similarly ignores which fingers are used, and instead emphasizes hitting the correct keys as fast as possible. The goal of this study is to develop a learning support system for touch typing based on fingering detected when typing, and to evaluate the possibility of teaching touch typing to non-touch typists through practicing with this system. In this study, we propose a method to determine which fingers are used to press which keys by using video of fingering made when typing and data of the pressed keys. Furthermore, we conduct evaluation experiments with the system using this method and evaluate the effectiveness of the system. According to the experiment results, estimation accuracy tended to decrease as CPM values increase. We learned that there were cases in which incorrect estimations were made, and we classified the causes into two categories: errors caused by occlusion and errors due to the image sampling being unable to keep up with the fingering. It is evident from the experimental results and discussion that when this system is used for non-touch typists to learn touch typing, it gives correct estimation results and is effective. For further research, we suggest using this system to facilitate the learning of touch typing for non-touch typists. More specifically, learners should enter a sequence of letters shown on-screen. Messages should be displayed on-screen when an incorrect key is pressed or when the recommended finger is not used. We believe that touch typing skills can be easily acquired by making users aware of what fingers they use to press the keys.
Keywords
