Özet:
In this study, a new and novel method of developing both ergonomic and efficient keyboard layouts is introduced based on the touch typing principles. A multi objective mathematical model with two objectives is defined. One of the objectives is related with the speed of typing while the other is related with the distribution of typing load among the fingers according to their capabilities. The parameters of the model are desired target finger loads and digraph costs; and monograph and digraph frequencies of the letters in the specified language. The experimentally determined penalty parameters are universal that can be used in keyboard layout optimization of any language. An optimization technique, namely “non-dominated sorting genetic algorithm II (NSGA II)” with Pareto dominance approach, is used for the application of the model and developing an ergonomic and optimal keyboard layout for Turkish language. Based on the optimization results, three of the obtained 152 layouts that minimize both monograph and digraph penalties are selected as the candidates of the best keyboard layout for Turkish language. These selected layouts were compared with F and Q layouts through optimization results, Dvorak’s criteria and experimental validation studies considering tendon travel of fingers and typing speed. The results indicate that all three candidate layouts are better than the Q and F layouts according to the optimization results and Dvorak’s criteria. With respect to the experimental validation studies, total tendon travel of three candidate layouts were less than F and Q layouts. In addition, tendon travels of two candidate layouts were relatively evenly distributed across fingers compared to the Q and F layouts. For the typing speed criterion, candidate layouts were better than Q layout, but not significantly different than F layout. Based on the overall results, ‘E layout’ which is one of the candidate layouts is selected as the best layout for touch typing in Turkish language.
