Number: 20060085757
Inventors: Andre; Bartley K.; (Menlo Park, CA) ; Ording; Bas; (San Francisco, CA) ; Christie; Greg; (San Jose, CA)
Filed: September 16, 2005
Virtual keyboard for use in “a personal computer, a computing system embedded within an electronic device (such as, for example, a cellular telephone, media player or other consumer electronic device.)”

Correlating what key a user is trying to hit and the sensory data from a touch screen can be difficult. In this application Apple suggests having different visual keys and sensory keys. They also suggest the use of contextual analysis and dictionary lookups to figure out what key is the most likely in a given context.
The average of the touch data [20] may not represent what the user understands where their click is directed [10]
It is desirable to process touches on a touch screen in a way that “makes sense,” which may include considering factors beyond (or instead of) a correspondence of the visual target and the touch area of a touch to activate the GUI to which the visual target corresponds."
Complex weighting of the likelihood of a key-press is directed at a particular key is done for each key and varies in response to context:
The weighting factor for each virtual key affects the size of that virtual key. The smaller the weighting factor associated with a first virtual key relative to the weighting factor associated with a second virtual key, the smaller the first virtual key is relative to the second virtual key (that is, the closer a touch location must be to a key location, relatively, to cause activation of the virtual key to which the key location corresponds.) In some examples, the weighting factors are normalized such that a weighting factor of one has no effect on the size of a virtual key, while a weighting factor greater than one has the effect of enlarging a virtual key, while a weighting factor less than one has the effect of diminishing the virtual key. The application of the weighting factors need not be linear. For example, the square root of the distance may be multiplied by the weighting factor.
In addition, the weights associated with particular virtual keys (see step 202 in FIG. 2 and step 606 in FIG. 6) may be adjusted dynamically. In one example, the weights are adjusted based on the sequence of virtual keys already activated. For example, the weight associated with the most likely next virtual key to be activated, based on the sequence of virtual keys already activated, may be set to be higher (or lower) than the weights associated with other virtual keys. What virtual key is most like to be activated may be determined, for example, using a dictionary (word-based), statistical probability for individual letters (for example, based on statistical frequency of occurrence), or a combination of both.
‘Dangerous’ keys are made virtually smaller
In some examples, virtual keys deemed to be “dangerous” (e.g., whose the result of whose activation may be difficult to reverse) are given a smaller weight. Potential examples of such “dangerous” keys may include, for example, a “delete” key or “cancel” key (for example, the “delete” GUI item 906 in FIG. 9.) Furthermore, in some examples, the slower virtual keys are activated (corresponding, nominally, to more deliberate activation), the less any particular virtual key would be weighted.
No Comments
Please Wait
Leave a Reply