UniTap - Comparison Chart

 

Characteristics presented in the table above are the following:

• Hands-free - number of hands (1 or 2) required for proper operation
• Portability - size of input surface, ability to be used in various devices
• Cost - cost of production (cost of operation)
• Speed - average speed of character input
• Accessibility - disability use
• Training - amount of training and adaptation required
• Flexibility - possible usage restrictions
• Simplicity - easiness in understanding and usage
• Language - language independence
• Privacy - level of privacy violation
• Density - number of available functions per finger area unit

 

Existing technologies and solutions for alphanumeric (word) input into compact devices include:

Voice recognition. It is software technology that processes voice input and tries to recognize known words said by the speaker. Its implementation has comparatively high processing power and memory requirements. User/software adaptation and learning is requested. It also implies out loud speaking, which might violate user's privacy.

Stroke recognition. This technology uses sensor elements (usually touch-sensitive screens of reasonably high resolution) which track path made by pen or stylus while writing one character. User thus writes one character in a time normally in quite friendly form, which, however, should be preliminary trained for better performance. Examples include UniStroke, Graffiti, etc.

Handwriting recognition. This technique is designed to process the complete path of a pen while writing the whole word in a regular handwriting manner. Although such solution would be extremely useful and convenient, reconstruction of the word from a stroke is a very difficult task not yet completely and efficiently solved.

On-screen keyboards. Devices having touch-sensitive screens, such as PDAs, may have an option of placing keyboard in a special area on a screen when alphanumeric input is required, e.g. email editing. This solution works well, but it consumes significant area of screen making the working area less readable. Besides, it can not be used on devices which do not have touch-sensitive panels.

12-button multitapping. This approach is currently the most widely-used one. According to this method, each button is assigned several characters which are selected cyclically depending how many times this button is pressed. Thus, typing a character requires about 2 finger strokes in average.

12-button predictive algorithms. This approach is also based on 12 buttons and character assignment similar to multitapping; however, it has dictionary of words stored in memory. Thus, user press sequence of buttons (one finger stroke per button) and the software suggests possible combinations of characters (words) from the dictionary. Examples of this approach include T9 (tm), iTap (tm), eZiTap (tm), etc.

12-button prefix-based algorithms. This approach is based on the prediction of a letter using already entered letters and language statistical database. This database is smaller then dictionaries of word-predictive algorithms. It's a good combination of multitapping simplicity and speed of predictive methods. Examples of this approach include LetterWise (tm) and WordWise (tm). WordWise (tm) requires 2 hands for input.

Integrated QWERTY keyboards. This technology stands for regular keyboard having one button per each character integrated into mobile device. Significant reduction in button size makes such keyboards practically inconvenient as several buttons tend to be pressed by a single finger stroke. Examples of such solution include Treo (tm), Psion, Nokia 9xxx Communicators, etc.

Accessory QWERTY keyboards. These are miniature keyboards which are provided as accessory to the devices having other means of character input (usually, 12-button multitapping). Examples include Sony Ericsson Chatboard (tm), Motorola iBoard (tm), Cirque Pocket Keyboard, etc.

Fastap (tm). According to this technique, character buttons are placed in the space between 12 conventional buttons representing digits in the telephone keypad. They also differ in size and elevation which make them distinguishable by the feel. Each character button has its own legend and should be pressed with no adjacent buttons touched in order to produce meaningful result. Details may be found at Digit Wireless.

UniTap. The UniTap technology allows for significant increase of functional elements density due to interpreting activation of several buttons as a meaningful character. However, in spite of the small size it provides ergonomic properties comparable to the regular-size keyboard and allows for uniform legend placement where all functional elements have equal elevation and shape.

 

This comparison chart is an expression of expert's view and may contain subjective classification of existing approaches.