Patent Grant
7953597
1. Field of the Invention
The present invention relates to multimodal browsers and voice servers and, more particularly, to voice-enabled multimodal applications for multimodal browsers and voice servers.
2. Description of the Related Art
Recent developments, many founded on the data-description eXtensible Markup Language (XML), have given rise to new Web-based applications including multimodal interfaces or browsers. A multimodal browser allows a user to access multimodal content, content that can be both graphical and audible. Traditionally, the user accessed Web content utilizing graphic input from a keyboard or manually directed screen-pointer entry. Later, the user also was been able to utilize speech input. More recently, the user has been able to access Web content through multimodal interfaces which permit the use of both graphic and speech inputs.
One type of multimodal browser is provided by the eXtensible Hypertext Markup Language (XHTML or XML)+Voice eXtensible markup language (VXML), also denoted more succinctly as the X+V markup language. The X+V markup language extends the traditional graphic browser to include spoken interactions. The X+V markup language integrates XHTML and XML-events technologies with XML vocabularies that were developed as part of the World Wide Web Consortium (W3C) Speech Interface Framework. The integration includes voice modules that support speech synthesis, speech dialogs, command and control applications, and speech grammars. Voice handlers can be attached to XHTML elements and respond to specific Document Object Model (DOM) events of a visual browser.
Notwithstanding these developments, a number of user desirable capabilities are not present in conventionally implemented multimodal interfaces, such as a user-friendly capability to fill form fields based upon speech utterances. Forms requiring user input have become commonplace. For example, users must commonly complete a form before being granted access privileges to enter a secure Web site. Inputting form information can be tedious, time consuming, and even frustrating. This can be especially true for a user who repetitively accesses content from various Web sites, each of which requires form-based input of user data before access is allowed. Moreover, the user may be using a device to access Web content that has limited or inconvenient input options. For example, a telephone, mobile phone, personal digital assistant (PDA), or similar type device often includes only a limited array of keys, a very small keypad, or nothing other than a voice input mechanism. It is desirable, therefore, that multimodal browsers be extended to provide an efficient way of voice enabling the automatic filling of form fields.
The present invention provides a computer-implemented method of automatically filling a form field in response to a speech utterance. The method can include the step of generating at least one grammar corresponding to the form field. The grammar can be based on a user profile and can include a semantic interpretation string. The method further can include the step of creating an event. The event can be based upon the at least one grammar and can be responsive to the speech utterance. The event can cause the filling of the form field with data corresponding to the user profile.
According to another embodiment, the present invention provides a computer system for automatically filling a form field in response to a speech utterance. The system can include a grammar-generating module that generates at least one grammar corresponding to the form field. The grammar can be based on a user profile and can comprise a semantic interpretation string. The computer system also can include an event module that creates an event that is based upon the at least one grammar and that is responsive to the speech utterance. The event can induce the filling of the form field with data corresponding to the user profile.
There are shown in the drawings, embodiments that are presently preferred; it being understood, however, that the invention is not limited to the precise arrangements and instrumentalities shown.
It will be readily apparent from the ensuing description that the illustrated multimodal communications environment 100 is but one type of multimodal communications environment in which the system 200 can be advantageously employed. Alternative multimodal communications environments, for example, can include various subsets of the different components illustratively shown.
Referring additionally to
The system 200 additionally includes a parsing module 217 that parses a document written in a VoiceXML and determines whether the document contains a synchronized voice field in a user profile domain. The term synchronized voice field is used herein to denote a form field which is filled by the synchronizing of speech and graphic inputs. As described below, this synchronization results in the filling of the form field with graphic input in response to a speech input. The term user profile domain is used herein to denote those form fields that are to be filled with data corresponding to a user profile, the user profile denoting for example, personal data corresponding to the user. Such personal information can include the user's name, address and phone number Other types of data alternately can be contained in the user profile and, accordingly, can be the subject of a voice-enabled auto-fill without changing the present invention as described herein.
The user profile in the present context illustratively includes a Key, Label Phrase, and Value, as shown by the representative user profile scheme in Table 1.
The API 210 provides a VoiceXML field grammar that implements a synchronization element, such as the X+V <sync> element. The synchronization element implemented by the VoiceXML field grammar synchronizes the value property of a graphic input control, such as an XHTML input control, with the synchronized voice field. As noted above, in the present context, the synchronized voice field defines the form field that is to be automatically filled, or auto-filled, by the system 200.
The grammar-generating module 215 can comprise a VoiceXML interpreter. As illustrated in
When one of the form fields so identified is accessed by a form interpretation algorithm (FIA), the grammar-generating module 215 enables the auto-fill grammar along with the VoiceXML field grammar provided by the API 210. When the auto-fill grammar and the VoiceXML field grammar are enabled, the event module 220 creates an auto-fill event based upon the grammars. The event is configured to respond to a speech utterance.
The auto-fill event, in response to the speech utterance, causes an execution of the SI string or tag so that the result of the execution is the value corresponding to the label. The auto-fill event causes the result to be propagated, and the synchronization element implemented by the VoiceXML field grammar fills in the form field of the Web page with the result of the SI string or tag execution. Accordingly, the event causes the form field to be filled with the value contained in the semantic interpretation string in response to the speech utterance.
The application program 305 can generate and pass a VoiceXML fragment 330 to the interpreter 312. The VoiceXML fragment 330 can specify a grammar that can be used to process received speech utterances. In case multiple devices are enabled for multimodal interaction, a configuration file 320 can be optionally included to specify one or more different devices such as a telephone, mobile phone, home security system, dashboard audio/communication system, computer system, and portable computer system. Within the configuration file 320, each device can be assigned an identifier that uniquely identifies that device. In one embodiment, prior to registering the VoiceXML fragment 330 with the interpreter 312, the application 305 can access the configuration file 320 to obtain the identity of the device being used.
The system 300 utilizes a command, control, and content navigation markup language (C3N), in which the application program 305 registers a VoiceXML link, based upon C3N grammars, such as a VoiceXML <link>, with the interpreter 312. Events generated by matching grammars in the link are propagated back to the application program 305 as shown. By specifying one or more link elements, based upon C3N grammars and denoted C3N link grammars, speech inputs to the application 305 can be matched. That is, the interpreter 312 can match speech inputs received from the application 305 with the C3N link grammars. Upon detecting a match, the interpreter 312 can generate one or more events that are sent back to the application 305.
More particularly, as illustrated in
For example, assuming the same user profile as set forth above, the following VoiceXML-based application uses the Key, Label Phrase, and Value elements to generate a grammar in accordance with this embodiment:
The grammar is built using the label phrases and corresponding keys, which are included in a portion of a corresponding SI string or tag. The grammar is illustratively constructed so as to match a phrase such as “fill my street address.” The VoiceXML link in response to the speech utterance causes an event to be propagated. The system 300 responds to the event, which is interpreted as an auto-fill command, by searching the user profile to obtain the value for the address. The result is that the form field is auto-filled with the value “8051 Congress Avenue.”
At step 414, an auto-fill grammar is generated based upon the label phrase and corresponding value of a user profile, the value being contained in an SI string or tag. The VoiceXML field grammar and the auto-fill grammars are enabled at step 416. In response to a speech utterance, the SI string is executed so that the result is the value contained in the SI string or tag at step 418. A visual field, at step 420, is auto-filled with the result.
As already noted, a system according to the different embodiments described herein can operate in different capacities, including functioning as a browser having an auto-fill capability for filling the fields of a Web form with graphical content in response to a speech utterance. The auto-fill capability can be especially beneficial if the browser resides on a device with a small form factor, such as a mobile phone.
The multimodal auto-fill, as described above, typically employs a grammar based upon user information that is supplied to the browser. The user identification, as also already noted, can comprise a personal profile or other identifying indicia, as well as any other type of information that the user can conveniently enter via the multimodal auto-fill supported by a particular browser. When the user information on which such a grammar is based is to be with a device having a small form factor, however, the loading of the grammar on the device can be a difficult task. Moreover, some devices many not have adequate memory space for storing the information. A particular device also may not have the capability to support local speech recognition.
A way to overcome these obstacles is provided by another embodiment of the present invention. The particular embodiment permits the user of a browser to enter user information over a data communications network. Any type of data communications network can be employed ranging from a local area network (LAN), such as a wireless LAN (WLAN), to the Internet. The user information can be supplied via a personal computer (PC) or other computing device connected to the particular data communications network.
More particularly, the user information can be maintained in a user information dialog application that is stored at a predetermined location of a particular data communications network, such as at a secure Web site on the Internet. The dialog generates grammars corresponding to the user information. For example, the grammars can be stored on a Web site as a set of files in a standard format such as the Speech Recognition Grammar Specification (SRGS).
If the user's device supports local speech recognition, then the multimodal auto-fill grammars are downloaded to the device after the user logs onto the information dialog application using the device's browser. Later when, for example, the user accesses a Web form provided by another application requiring user information, the user fills in the form using the multimodal auto fill capability as described above.
The dialog application 600 illustratively resides on a Web server 608. The user provides the user-specific information to the dialog application 600 from a client 610, such as a desktop Web client, via a data communications network connection between the client and the Web server 608. Subsequently, using any device that has speech recognition capability as well as auto-fill capability, the user links to the dialog application 600 using the device's browser and downloads from the grammar storage module 604 the grammars that are stored thereon. When the user later accesses a Web form provided by some other application that requests user-specific information, the user fills in the form using the multimodal capability.
Table 2 is illustrative of user-specific information that can comprise a user profile.
œfirstnameâ
□
œfirst nameâ
□
œSoonthornâ
□
œlastnameâ
□
œlast nameâ
□
œAtivanichayaphongâ
□
œaddressâ
□
œstreet
œ8051 Congress Aveâ
□
□
œphoneâ
□
œphone
œ561-666-1234â
□
□
The following exemplary grammars can be generated by the grammar generation module 602 for the user-specific information in Table 2. Note that the format is the Java API Speech Grammar Format (JSGF) for representing speech grammars, but other grammars such as the W3C Speech Recognition Grammar Specification can alternately be used.
For a user device having a speech recognition capability, the above grammars are fetched after the user logs onto the web application using the device's web browser. These exemplary grammars can be stored by the Web application as separate grammar files that can be accessed using the following exemplary URL's:
An alternative embodiment of a user-information dialog application is schematically illustrated in
For example, the address can correspond to URL locations that are included with an HTTP request for a Web page, the request coming from the browser operating on the user device 902. More particularly, the request can contain an HTTP 1.1 content-location header, as will be readily understood by one of ordinary skill in the art. The multimodal application 904 receives the HTTP request and reads the URL contained within the content-location headers. The multimodal application 904 then fetches the grammars both generated and stored by the dialog application 900, and then supplies the grammars to a speech recognition engine (not explicitly shown) located on the application server 906. The multimodal auto-fill grammars are activated by the speech engine so that the grammars can be used in conjunction with the Web form response sent by the application to the user.
Note that for a user device that does not support local speech recognition, the URL's are downloaded after the user logs onto a web application using the Web browser running on the device. When the user later uses the Web browser to access a particular Web form provided by a remote multimodal application, the URL's are included with the HTTP 1.1 request for the Web page. The proposed HTTP 1.1 header to be used is content-location:
The multimodal Web application reads the content-header information and fetches the grammars. Note further that, as alluded to above, the application server can use an authentication mechanism, which may require the client's tokens (i.e., “cookies”) to be passed to the Web server storing the grammars. The tokens could contain the IP address of the user's small device, which is unique.
At step 1008, the application presents a dialog for entering user-specific information, such as the user's name, address, and/or telephone number. The application associates the user-supplied information with the IP address of the user device and stores the associated data in a database at step 1010. The user can then log off from the application at step 1012.
A determination is illustratively made at step 1014 as to whether the particular device corresponding to the now-registered information has a speech recognition capability. If it does, then the auto-fill grammars are associated with the user-specific information are downloaded to the device at step 1016. Otherwise, at step 1018, the location on the Web site where the associated auto-fill grammars reside are downloaded to the device. The method illustratively concludes at step 1020.
If, however, the device does support speech recognition, then steps 1106-1110 can be circumvented. Regardless, the method continues at step 1112 with the Web application presenting an XHTML form and, optionally, a voice dialog for filling in the form. The grammars for filling in the form are added at step 1114, and various methods described above for performing an auto-fill are subsequently formed at step 1116. The method concludes at step 1118.
The present invention can be realized in hardware, software, or a combination of hardware and software. The present invention can be realized in a centralized fashion in one computer system, or in a distributed fashion where different elements are spread across several interconnected computer systems. Any kind of computer system or other apparatus adapted for carrying out the methods described herein is suited. A typical combination of hardware and software can be a general purpose computer system with a computer program that, when being loaded and executed, controls the computer system such that it carries out the methods described herein.
The present invention also can be embedded in a computer program product, which comprises all the features enabling the implementation of the methods described herein, and which when loaded in a computer system is able to carry out these methods. Computer program in the present context means any expression, in any language, code or notation, of a set of instructions intended to cause a system having an information processing capability to perform a particular function either directly or after either or both of the following: a) conversion to another language, code or notation; b) reproduction in a different material form.
This invention can be embodied in other forms without departing from the spirit or essential attributes thereof. Accordingly, reference should be made to the following claims, rather than to the foregoing specification, as indicating the scope of the invention.
The present continuation-in-part application claims the benefit of U.S. patent application Ser. No. 10/945,112 filed Sep. 20, 2004, which is incorporated in its entirety herein by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5619708 | Ho | Apr 1997 | A |
| 5642519 | Martin | Jun 1997 | A |
| 5995918 | Kendall et al. | Nov 1999 | A |
| 6058366 | Tarkiainen et al. | May 2000 | A |
| 6078886 | Dragosh et al. | Jun 2000 | A |
| 6188985 | Thrift et al. | Feb 2001 | B1 |
| 6199079 | Gupta et al. | Mar 2001 | B1 |
| 6240448 | Imielinski et al. | May 2001 | B1 |
| 6314402 | Monaco et al. | Nov 2001 | B1 |
| 6385583 | Ladd et al. | May 2002 | B1 |
| 6456974 | Baker et al. | Sep 2002 | B1 |
| 6490601 | Markus et al. | Dec 2002 | B1 |
| 6658414 | Bryan et al. | Dec 2003 | B2 |
| 6856960 | Dragosh et al. | Feb 2005 | B1 |
| 6898567 | Balasuriya | May 2005 | B2 |
| 6975993 | Keiller | Dec 2005 | B1 |
| 7003464 | Ferrans et al. | Feb 2006 | B2 |
| 7050977 | Bennett | May 2006 | B1 |
| 7076428 | Anastasakos et al. | Jul 2006 | B2 |
| 7146323 | Guenther et al. | Dec 2006 | B2 |
| 7167824 | Kallulli | Jan 2007 | B2 |
| 7177814 | Gong et al. | Feb 2007 | B2 |
| 7200559 | Wang | Apr 2007 | B2 |
| 7216351 | Maes | May 2007 | B1 |
| 7260535 | Galanes et al. | Aug 2007 | B2 |
| 7382770 | Bergman et al. | Jun 2008 | B2 |
| 20020054090 | Silva et al. | May 2002 | A1 |
| 20020059073 | Zondervan et al. | May 2002 | A1 |
| 20020091520 | Endo et al. | Jul 2002 | A1 |
| 20020093530 | Krothapalli et al. | Jul 2002 | A1 |
| 20020107918 | Shaffer et al. | Aug 2002 | A1 |
| 20020165719 | Wang et al. | Nov 2002 | A1 |
| 20020193990 | Komatsu | Dec 2002 | A1 |
| 20020198719 | Gergic et al. | Dec 2002 | A1 |
| 20030028792 | Plow et al. | Feb 2003 | A1 |
| 20030033146 | Morin et al. | Feb 2003 | A1 |
| 20030088421 | Maes et al. | May 2003 | A1 |
| 20030115060 | Junqua et al. | Jun 2003 | A1 |
| 20030140113 | Balasuriya | Jul 2003 | A1 |
| 20030153362 | Goldsmith et al. | Aug 2003 | A1 |
| 20030182622 | Sibal et al. | Sep 2003 | A1 |
| 20040113908 | Galanes et al. | Jun 2004 | A1 |
| 20040128135 | Anastasakos et al. | Jul 2004 | A1 |
| 20040138890 | Ferrans et al. | Jul 2004 | A1 |
| 20040141011 | Smethers et al. | Jul 2004 | A1 |
| 20040243393 | Wang | Dec 2004 | A1 |
| 20050004800 | Wang | Jan 2005 | A1 |
| 20050028085 | Irwin et al. | Feb 2005 | A1 |
| 20050171762 | Ryan et al. | Aug 2005 | A1 |
| 20060167686 | Kahn | Jul 2006 | A1 |
| 20060265225 | Ichnowski | Nov 2006 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20060074652 A1 | Apr 2006 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10945112 | Sep 2004 | US |
| Child | 11199672 | US |
