Modeling Human Performance of Stroke-Based Text Entry

Gesture typing is a popular stroke-based text input method for mobile devices. Instead of tapping, the user enters a word with a single continuous stroke by gesturing through all the letters. In this paper, we introduce a quantitative motor control model to predict the human performance of gesture typing in terms of production times. Based on a model of cursive handwriting, it can be applied to predict the influence of the keyboard layout on gesture typing performance. In contrast to previous approaches, the "chunking" of multiple movements into a single action, which is observed for expert users, is a natural characteristic of the proposed model that avoids over-estimation of the production times. Data collected from gesture keyboard users is evaluated to assess the performance of the model in comparison to previous models.