Original Publication Date: 2003-Jun-13
Included in the Prior Art Database: 2003-Jun-13
A means for leveraging a web browser's Autofill feature in order to provide increase speech recognition accuracy when speech is used in multimodal interfaces.
Electronic forms require users to fill in fields and answer questions. Web browsers provide an "Autofill" feature that allows the user to specify commonly used information like Name, Address, phone, etc., to make ifilling out that information on a form easier. Basically, the browser recognizes a field as prompting for the information in its Autofill feature. When the starts typing into the field, the browser finishes filling it based on the information the user provided in the Autofill preference.
Multimodal interfaces add speech recognition and synthesis to the interface. For mobile devices that lack a keyboard or require a free hand for a stylus, speech recognition is a very efficient way to provide input. However, saying your city name is different that spelling it. This means that fields in a browser's Autofill feature are not activated for use in speech interaction.
A means for leveraging the accuracy and speed of a browser's Autofill feature needs to be provided in multimodal interfaces.
This invention solves this problem by automatically activating built-in spelling and digit grammars (as appropriate) for a field identified by a browser as being an Autofill field. Alternatively, the user may provide a word or part of a word as a grammar to substitute for the words intended to fill in a field. The autofill grammar is in addition to the application provided grammar for that field. While a speech utterance to completely fill the field may...