Improving dictionary-based disambiguation text entry method accuracy

Text entry on mobile devices is problematic because of ever-decreasing device sizes. Dictionary-based keypad text entry methods are relatively effective, but still run into problems of word ambiguity, especially when used with small numbers of keys. Common text entry disambiguation methods only use word frequency information to resolve conflicts. This paper proposes a new method that also looks at semantic information (distances between word meanings). Simulations show encouraging results, suggesting potential practical applications of this method to mobile devices.