Optimal character arrangement for ambiguous keyboards using a PSO-based algorithm

Many researchers have devoted their efforts on optimizing the character arrangement for the single-character keyboard (SCK), however, the SCK market is still dominated by the QWERTY keyboard because people are used to the typology and hesitate to learn a new one even the QWERTY keyboard is ill-designed. As the use of hand-held entry devices is growing, the optimal arrangement of multi-character keyboard (MCK) is increasingly important. This also apples to customized keyboard design for motor-impaired users who are usually not able to do the typing on many keys. In this paper, we propose a unified mathematical model that integrates varying objectives. An enhanced particle swarm algorithm is developed by embedding a character-swapping optimizer and a bounding technique. Experimental results manifest that the proposed method outperforms competing ones by taking into account the co-occurrence frequency of characters in words, typing ergonomics, and word-disambiguation effectiveness.