Speech-based text entry for mobile handheld devices: An analysis of efficacy and error correction techniques for server-based solutions

As handheld devices become ubiquitous and the tasks performed become multipurpose in nature, efficient data entry techniques are necessary. This research evaluated several speech-based text entry solutions for handheld devices using server-based speech recognition. Because server-based solutions introduce network delays, an analysis of the relationship among network delays, number of recognition errors, how fast users can correct errors, and overall data entry rates was performed. The analysis and empirical results confirm the importance of minimizing recognition errors. This suggests that a server-based approach that makes more computing resources available may prove effective. Results from two empirical studies are presented. The first compares two error correction mechanisms: a multitap and soft keyboard solution. The second employs a longitudinal investigation of the effects of experience on text entry rates. Users attained an effective mean text entry rate as high as 25.3 words per min, which is higher than or comparable to data entry rates reported for other input techniques for handheld devices. The results of this research have implications for researchers and designers of automatic speech recognition systems and mobile devices.