Design Dimensions of Intelligent Text Entry Tutors

Intelligent text entry methods use techniques from artificial intelligence to improve entry rates. While these text entry methods are useful in situations when a full-sized keyboard is impractical or unavailable, they also require substantial training investment from users. We hypothesize that intelligent text entry tutors may reduce this time and effort. However, before we set out to design these tutors we need to consider their design space. This paper contributes to this understanding by proposing and analyzing five design dimensions: automaticity, error correction, coverage, feedback and engagement.