User profiling for voice input processing

[0001] This application is a continuation of U.S. patent application Ser. No. 12/712,988, filed Feb. 25, 2010, which is incorporated herein by reference for all purposes.

[0002] This is directed to processing received voice inputs by identifying an instruction likely to be provided by the user of the voice. In particular, this is directed to identifying the user providing a voice input and processing the voice input using a subset of resources

[0003] Many electronic devices provide a significant number of features or operations accessible to a user. The number of available features or operations may often exceed the number of inputs available using an input interface of the electronic device. To allow users to access electronic device operations that are not specifically tied to particular inputs (e.g., inputs not associated with a key sequence or button press, such as a MENU button on an iPod, available from Apple Inc.), the electronic device may provide menus with selectable options, where the options are associated with electronic device operations. For example, an electronic device may display a menu with selectable options on a display, for example in response to receiving an input associated with the menu from an input interface (e.g., a MENU button).

[0004] Because the menu is typically displayed on an electronic device display, a user may be required to look at the display to select a particular option. This may sometimes not be desirable. For example, if a user desires to conserve power (e.g., in a portable electronic device), requiring the electronic device to display a menu and move a highlight region navigated by the user to provide a selection may use up power. As another example, if a user is in a dark environment and the display does not include back lighting, the user may not be able to distinguish displayed options of the menu. As still another example, if a user is blind or visually impaired, the user may not be able to view a displayed menu.

[0005] To overcome this issue, some systems may allow users to provide instructions by voice. In particular, the electronic device can include audio input circuitry for detecting words spoken by a user. Processing circuitry of the device can then process the words to identify a corresponding instruction to the electronic device, and execute the corresponding instruction. To process received voice inputs, the electronic device can include a library of words to which the device can compare the received voice input, and from which the device can extract the corresponding instruction.

[0006] In some cases, however, the size of the word library can be so large that it may be prohibitive to process voice inputs, and in particular time and resource-prohibitive to process long voice inputs. In addition, the electronic device can require significant resources to parse complex instructions that include several variables provided as part of the voice instruction (e.g., an instruction that includes several filter values for selecting a subset of media items available for playback by the electronic device).

[0007] This is directed to systems and methods for identifying a user providing a voice input, and processing the input to identify a corresponding instruction based on the user's identity. In particular, this is directed to processing a received voice input using the subset of library terms used to process the voice input.

[0008] An electronic device can receive a voice input for directing the device to perform one or more operations. The device can then process the received input by comparing the analog input signal with words from a library. To reduce the load for processing the received voice input, the electronic device can limit the size of a library to which to compare the voice input (e.g., the number of library words) based on the identity of the user providing the input.

[0009] The electronic device can identify the user using any suitable approach. For example, the electronic device can identify a user from the content of an input provided by the user (e.g., a user name and password). As another example, the electronic device can identify a user by the type of interaction of the user with the device (e.g., the particular operations the user directs the device to perform). As still another example, the electronic device can identify a user based on biometric information (e.g., a voice print). Once the user has been identified, the electronic device can determine the user's interests and define the library subset based on those interests. For example, the subset can include words corresponding to metadata related to content selected by the user for storage on the device (e.g., transferred media items) or content added to the device by the user (e.g., the content of messages sent by the user). As another example, the subset can include words corresponding to application operations that the user is likely to use (e.g., words relating to media playback instructions).

[0010] In response to identifying the words of a particular voice input, the electronic device can identify one or more instructions that correspond to the voice input. The instructions can then be passed on to appropriate circuitry of the electronic device for the device to perform an operation corresponding to the instruction. In some embodiments, the instruction can identify a particular device operation and a variable or argument characterizing the operation.

[0022] An electronic device is operative to receive voice inputs provided by a user to control electronic device operations. In particular, an electronic device is operative to receive and process voice inputs to identify words spoken by the user, and to determine an instruction for performing a device operation corresponding to the identified words.

[0023] The electronic device can include a processor and an input interface that includes audio input circuitry. Using the audio input circuitry, a user can provide voice inputs to the device for directing the device to perform one or more operations. The voice inputs can have any suitable form, including for example pre-defined strings corresponding to specific instructions (e.g., “play artist Mika”), arbitrary or natural language instructions (e.g., “pick something good”), or combinations of these.

[0024] The electronic device can parse a received voice input to identify the words of the input. In particular, the electronic device can compare words of the received input with a library of words. In the context of an electronic device used to play back media items, the number of words in the library can be significant (e.g., including the artist names, album names and track names of media items in a user's media library). Comparing the voice input to an entire word library can take a significant amount of time, so it may be beneficial to reduce the amount of the library to which the voice input is compared. In some embodiments, one or more subsets can be defined in the voice library based on the identity of the user providing the voice input.

[0025] The electronic device can define, for each user, a preference profile or other information describing the users interests, the particular manner in which the user typically interacts with the device, or both. For example, the profile can include information identifying the types of media items played back by the user, applications used by the user, typical playback behavior (e.g., pick a playlist and don't interact much with the device, or regularly change the played back media item). As another example, the profile can include information regarding the types of media items that the user typically plays back or does not play back. Using the profile information, the electronic device can define a subset of library words that relate to the profile, and initially limit or reduce the processing of a received voice command to the defined subset of library words.

[0026] The electronic device can identify the user using any suitable approach. In some embodiments, the electronic device can identify the user based on a particular input of the user (e.g., the entry of a username or password), from attributes of the entry (e.g., a voice print of the voice input), biometric information detected by the device, or any other suitable approach.

[0027] FIG. 1 is a schematic view of a electronic device in accordance with one embodiment of the invention. Electronic device 100 may include processor 102, storage 104, memory 106, input interface 108, and output interface 110. In some embodiments, one or more of electronic device components 100 may be combined or omitted (e.g., combine storage 104 and memory 106). In some embodiments, electronic device 100 may include other components not combined or included in those shown in FIG. 1 (e.g., communications circuitry, location circuitry, sensing circuitry detecting the device environment, a power supply, or a bus), or several instances of the components shown in FIG. 1. For the sake of simplicity, only one of each of the components is shown in FIG. 1.

[0028] Processor 102 may include any processing circuitry or control circuitry operative to control the operations and performance of electronic device 100. For example, processor 102 may be used to run operating system applications, firmware applications, media playback applications, media editing applications, or any other application. In some embodiments, a processor may drive a display and process inputs received from a user interface.

[0029] Storage 104 may include, for example, one or more storage mediums including a hard-drive, solid state drive, flash memory, permanent memory such as ROM, any other suitable type of storage component, or any combination thereof. Storage 104 may store, for example, media data (e.g., music and video files), application data (e.g., for implementing functions on device 100), firmware, user preference information (e.g., media playback preferences), authentication information (e.g. libraries of data associated with authorized users), lifestyle information (e.g., food preferences), exercise information (e.g., information obtained by exercise monitoring equipment), transaction information (e.g., information such as credit card information), wireless connection information (e.g., information that may enable electronic device 100 to establish a wireless connection), subscription information (e.g., information that keeps track of podcasts or television shows or other media a user subscribes to), contact information (e.g., telephone numbers and email addresses), calendar information, and any other suitable data or any combination thereof.

[0030] Memory 106 can include cache memory, semi-permanent memory such as RAM, and/or one or more different types of memory used for temporarily storing data. In some embodiments, memory 106 can also be used for storing data used to operate electronic device applications, or any other type of data that may be stored in storage 104. In some embodiments, memory 106 and storage 104 may be combined as a single storage medium.

[0031] Input interface 108 may provide inputs to input/output circuitry of the electronic device. Input interface 108 may include any suitable input interface, such as for example, a button, keypad, dial, a click wheel, or a touch screen. In some embodiments, electronic device 100 may include a capacitive sensing mechanism, or a multi-touch capacitive sensing mechanism. In some embodiments, input interface can include a microphone or other audio input interface for receiving a user's voice inputs. The input interface can include an analog to digital converter for converting received analog signals corresponding to a voice input to a digital signal that can be processed and analyzed to identify specific words or instructions.

[0032] Output interface 110 may include one or more interfaces for providing an audio output, visual output, or other type of output (e.g., odor, taste or haptic output). For example, output interface 110 can include one or more speakers (e.g., mono or stereo speakers) built into electronic device 100, or an audio connector (e.g., an audio jack or an appropriate Bluetooth connection) operative to be coupled to an audio output mechanism. Output interface 110 may be operative to provide audio data using a wired or wireless connection to a headset, headphones or earbuds. As another example, output interface 110 can include display circuitry (e.g., a screen or projection system) for providing a display visible to the user. The display can include a screen (e.g., an LCD screen) that is incorporated in electronic device 100, a movable display or a projecting system for providing a display of content on a surface remote from electronic device 100 (e.g., a video projector), or any other suitable display. Output interface 110 can interface with the input/output circuitry (not shown) to provide outputs to a user of the device.

[0033] In some embodiments, electronic device 100 may include a bus operative to provide a data transfer path for transferring data to, from, or between control processor 102, storage 104, memory 106, input interface 108, output interface 110, and any other component included in the electronic device.

[0034] A user can interact with the electronic device using any suitable approach. In some embodiments, the user can provide inputs using one or more fingers touching an input interface, such as a keyboard, button, mouse, or touch-sensitive surface. In some embodiments, a user can instead or in addition provide an input by shaking or moving the electronic device in a particular manner (e.g., such that a motion sensing component of the input interface detects the user movement). In some embodiments, a user can instead or in addition provide a voice input to the electronic device. For example, the user can speak into a microphone embedded in or connected to the electronic device.

[0035] The user can provide voice inputs to the electronic device at any suitable time. In some embodiments, the electronic device can continuously monitor for voice inputs (e.g., when the device is not in sleep mode, or at all times). In some embodiments, the electronic device can monitor for voice inputs in response to a user input or instruction to enter a voice input. For example, a user can select a button or option, or place the electronic device in such a manner that a sensor detects that the user wishes to provided a voice input (e.g., a proximity sensor detects that the user has brought the device up to the user's mouth). In some embodiments, the electronic device can monitor for user inputs when one or more particular applications or processes are running on the device. For example, the electronic device can monitor for voice inputs in a media playback application, a voice control application, a searching application, or any other suitable application.

[0036] FIG. 2 is a schematic view of an illustrative display dedicated to monitoring for a voice input in accordance with one embodiment of the invention. Display 200 can include information region 210 indicating that the device is monitoring for a voice input. For example, information region 210 can include title 212 specifying the name of the application or process monitoring for a voice input. As another example, information region 210 can include waveform 214 providing a depiction of the detected voice input. The content displayed in information region 210 can change dynamically as a received input is detected (e.g., wave form 214 changes), or the content can instead or in addition remain static. Display 200 can include option 220 for directing the device to initialize or cancel monitoring for a voice input. For example, option 220 can switch between “start” and “cancel” options based on the state of the device. In some cases, the electronic device can instead or in addition include a “complete” option for indicating when a voice input has been completely entered.

[0037] In some embodiments, the electronic device can display one or more discreet elements on an existing electronic device display to indicate that the device is monitoring for voice inputs. FIG. 3 is a schematic view of an illustrative display having an indication that the device is monitoring for voice inputs in accordance with one embodiment of the invention. Display 300 can include a “now playing” display for a media playback application. For example, display 400 can include art 310 depicting a current, previous or future media item to be played back, and media selection options 320. To indicate that the electronic device is monitoring for voice inputs, display 300 can include element 330. Element 330 can include any suitable display element, including for example text, a graphic, glyph, or other content. In some embodiments, element 330 can include one or more elements of display 200 (FIG. 2) In some embodiments, element 330 can instead or in addition include a selectable option for enabling or disabling the monitoring for voice inputs. This may allow a user to control the resource consumption required for voice inputs. In some embodiments, element 330 can instead or in addition provide an indication that monitoring for voice inputs is available if the user provides a proper input using the input interface (e.g., the user approaches the device to his mouth, as detected by a proximity sensor of the device).

[0038] Voice inputs can include instructions for performing any suitable electronic device operation. In some embodiments, voice inputs can relate to a specific set or library of instructions that the device can detect. For example, the device can be limited to detecting particular keywords for related to specific device operations, such as “play,” “call,” “dial,” “shuffle,” “next,” “previous,” or other keywords. In some cases, each keyword can be accompanied by one or more variables or arguments qualifying the particular keyword. For example, the voice input can be “call John's cell phone,” in which the keyword “voice” is qualified by the phrase “John's cell phone,” which defines two variables for identifying the number to call (e.g., John and his cell phone). As another example, the voice input can be “play track 3 of 2005 album by the Plain White T's,” in which the keyword “play” is qualified by the phrase “track 3 of 2005 album by the Plain White T's.” This phrase has three variables for identifying a particular song to play back (e.g., artist Plain White T's, 2005 album, and track 3). As still another example, the phrase “shuffle then go next five times” can include two keywords, “shuffle” and “next” as well as a qualifier for the “next” keyword (e.g., “five times”).

[0039] In some cases, the electronic device can detect and parse natural language voice inputs. For example, the electronic device can parse and process an input such as “find my most played song with a 4-star rating and create a Genius playlist using it as a seed.” This voice input can require significant processing to first identify the particular media item to serve as a seed for a new playlist (e.g., most played song with a particular rating), and then determine the operation to perform based on that media item (e.g., create a playlist). As another example, a natural language voice input can include “pick a good song to add to a party mix.” This voice input can require identifying the device operation (e.g., add a song to a party mix) and finding an appropriate value or argument to provide the device operation, where the value can be user-specific.

[0040] The voice input provided to the electronic device can therefore be complex, and require significant processing to first identify the individual words of the input before extracting an instruction from the input and executing a corresponding device operation. The electronic device can identify particular words of the voice input using any suitable approach, including for example by comparing detected words of the voice input to a library or dictionary of locally stored words. The library can include any suitable words, including for example a set of default or standard words that relate generally to the electronic device, its processes and operations, and characteristics of information used the processes and operations of the device. For example, default words in the library can include terms relating to operations of one or more applications (e.g., play, pause, next, skip, call, hang up, go to, search for, start, turn off), terms related to information used by applications (e.g., star rating, genre, title, artist, album, name, play count, mobile phone, home phone, address, directions from, directions to), or other such words that may be used for by any user of an electronic device.

[0041] In some embodiments, the library can instead or in addition include words that relate specifically to a user of the device. For example, the library can include words determined from metadata values of content or information stored by the user on the device. Such words can include, for example, titles, artists and album names of media items stored by a user on the device, genre, year and star rating values for one or more media items, contact names, streets, cities and countries, email addresses, or any other content that a user can store on the device that may be specific to a particular user. The electronic device can define a library using any suitable approach, including for example by augmenting a default library with words derived from user-specific content of a user using the device.

[0042] FIG. 4 is a schematic view of an illustrative system for identifying device operations to perform in response to a voice input in accordance with one embodiment of the invention. System 400 can include voice input 410 provided by a user. Voice input 410 can be detected or received by any suitable combination of hardware, firmware and software for detecting audio provided by a user to an electronic device, converting the analog audio signal to a digital signal, and cleaning up the digital signal for further processing. For example, the electronic device can include a microphone for detecting the analog voice input, and an analog to digital converter for converting the voice input. The electronic device can encode the voice input using any suitable approach, including any suitable encoding scheme.

[0043] The voice input can be provided to voice input processing module 420. The provided voice input can be provided in any suitable form, including for example in digitized form or in analog form (e.g., if some or all of the circuitry and software for converting an analog voice input to a digital signal are in voice input processing module 420). For example, voice input processing module 420 can be integrated in the electronic device used by the user. As another example, voice input processing module can totally or in part be integrated in a remote device or server to which the device can connect to process voice inputs. Voice input processing module 420 can analyze the received voice input to identify specific words or phrases within the voice input. For example, voice input processing module 420 can compare identified words or phrases of the voice signal to words or phrases of library 422 of words. Library 422 can be separate from voice input processing module 420, or instead or in addition embedded within voice input processing module 420. Library 422 can include any suitable words, including for example default words associated with the electronic device detecting the voice input, specific words derived from the user's interactions with the electronic device (e.g., with content transferred to the electronic device by the user), or other words or phrases.

[0044] Voice input processing module 420 can analyze the detected words or phrases, and identify one or more particular electronic device operations associated with the detected words or phrases. For example, voice input processing module 420 can identify one or more keywords specifying an instruction to the device, where the instruction can include one or more variables or values qualifying the instruction. The instruction (e.g., “play”), including the variables or values specifying how the instruction is to be executed (e.g., “Mika's latest album”) can be analyzed to identify one or more electronic device operations corresponding to the instruction.

[0045] Voice input processing module 420 can provide the identified device operation to the device so that device 430 performs an operation. Device 430 can perform one or more operations, including for example operating one or more applications or processes within one or more applications, and can include a punctual, repeating, or lasting operation (e.g., monitor all incoming email for particular flagged messages). Device 430 can include any suitable device, and can include some or all of the features of electronic device 100 (FIG. 1). In some embodiments, device 430 can detect and provide voice input 410 to voice input processing module 420, which can also reside on device 430. In some embodiments, device 430 can instead or in addition be a distinct device that receives instructions to perform operations from a remote voice input processing module 420. Device 430 can receive instructions from processing module 420 over path 444. Processing module 420 can compare voice inputs received over path 440 with library 422 over path 442. Each of paths 440, 442 and 444 can be provided over any suitable communications network or protocol, including for example wired and wireless networks and protocols.

[0046] Because of the complexity of voice inputs, and the size of the resulting library used to identify instructions within a voice input, the voice input processing module can take a significant amount of time, resources, or both to process a particular voice input. To reduce the processing required for each voice input, the voice input processing module may benefit by comparing the voice input to a reduced set of library words. In particular, by reducing the number of words in the library to which a voice input is compared, the voice input processing module can more rapidly process voice inputs at a lower device resource cost.

[0047] The voice input processing module can determine which library words to include in a particular subset using any suitable approach. In some embodiments, a subset of the library can be selected based on the identity of the user providing the voice input. The voice input processing module can determine which words in a library to associate with a user using any suitable approach. For example, the voice input processing module can select default words that relate to applications or operations used often by the user (e.g., used more than a threshold amount). As another example, the voice input processing module can prompt the user to provide preference or interest information from which related library words can be extracted. As still another example, the voice input processing module can instead or in addition monitor the user's use of the device to determine the user's preferences. In some embodiments, the voice input processing module can analyze previously received voice inputs to identify particular words or types of words that are often used. FIG. 5 is a flowchart of an illustrative process for selecting a subset of a voice input library in accordance with one embodiment of the invention. Process 500 can begin at step 502. At step 504, a voice input processing module can determine whether a user providing a voice input has been identified. For example, the processing module can determine whether one or more characteristics of the current user match characteristics stored in memory. As another example, the processing module can determine whether the user has provided an input that is associated only with the user. If the processing module determines that the user has not been identified, process 500 can move to step 506 and end. Alternatively, the processing module can define a new profile for identifying the new user's interests, and move to step 508.

[0048] If, at step 504, the processing module instead determines that the user has been identified, process 500 can move to step 508. At step 508, the processing module can identify user interest information. In particular, the processing module can identify content or other information specifying the user's interests, and can use the information to generate a preference profile for the user. The processing module can identify user interest information using any suitable approach, including one or more of the approaches described within step 508. At step 510, the processing module can review past user use of the device. For example, the processing module can review feedback information related to media playback (e.g., which media items were selected for playback, skipped, or ranked). As another example, the processing module can review the particular applications or operations that the user directed the device to perform (e.g., the user often uses an email application and sports scores application). As still another example, the processing module can review the types of inputs that the user provided to particular applications or in the context of specific operations (e.g., the user is interested in baseball scores and news, but not basketball or hockey scores and news).

[0049] At step 512, the processing module can identify user-selected content stored on the device. For example, the processing module can identify attributes of media items that the user selected to transfer from a media library to the device. As another example, the processing module can identify attributes of particular applications that the user has installed or loaded on the device.

[0050] At step 514, the processing module can request preference information from the user. For example, the processing module can provide a number of questions to the user (e.g., select from the following list your preferred genres, or identify specific media items that you like). As another example, the processing module can direct the user to indicate a preference for currently provided content (e.g., direct the user to approve or reject a currently played back media item, or a game that the user is trying). At step 516, the processing module can review words identified from previous voice inputs. For example, the processing module can review previously received voice inputs, and the types of words or phrases identified in the previous inputs. In some embodiments, the processing module can further determine which of the identified words were properly identified (e.g., the words for which the corresponding device operation executed by the device was approved by the user).

[0051] At step 518, the processing module can identify particular library words associated with the user interest information. For example, the processing module can select a subset of default library words that are associated with particular operations or processes most often used by the user. As another example, the processing module can select a subset of user-specific library words that relate particularly to the content of most interest to the user (e.g., words for metadata related to the media items preferred by the user). In particular, the processing module can identify particular metadata associated with media items of most interest to the user (e.g., media items most recently added to the user's media library, transferred to the device, having the highest user ranking, popular media based on external popularity sources, media by a particular favorite artist or within a genre, media items with higher playcounts). At step 520, the processing module can define a subset of the library that includes at least the identified library words. In some embodiments, the defined subset can include additional words, including for example default library words, or other words commonly used or associated with other users (e.g., words associated with other users of the same device, with users using the same type of device, or with users within a particular community or location). Process 500 can then move to step 506 and end.

[0052] The voice input processing module can identify a user using any suitable approach. FIG. 6 is a flowchart of an illustrative process for identifying a user providing a voice input in accordance with one embodiment of the invention. Process 600 can begin at step 602. At step 604, the processing module can determine whether an input was received. For example, the processing module can determine whether an input interface of an electronic device has detected or received an input from a user. The input can be in any suitable form, including for example a voice input or an input provided by the user using his hand or fingers. If the processing module determines that no input has been received, process 600 can move to step 606 and end.

[0053] If, at step 604, the processing module instead determines that an input has been received, process 600 can move to step 608. At step 608, the processing module can identify the user providing the input. The processing module can identify user providing an input using any suitable approach, including one or more of the approaches described within step 608. At step 610, the processing module can identify the user from a user-specific input. For example, the processing circuitry can identify the user from a username and password, token, or other key or secret known only the user. At step 612, the processing module can identify the user from the type of input received. For example, the processing module can determine that the input corresponds to an operation or process typically performed by a particular user (e.g., only one user uses a particular application). As another example, the processing module can determine that the input was provided at a particular time of day during which the same user uses the device. As step 614, the processing module can identify the user from biometric information of the input. For example, the processing module can identify a user from a voiceprint, fingerprint, recognition of one or more facial features, or any other detected biometric attribute of the user (e.g., by comparing the biometric attribute to a library of known biometric attributes each associated with particular known users of the device).

[0054] At step 616, the processing module can use the user's identity for voice processing. In particular, the processing module can retrieve a subset of the word library used for processing voice inputs to streamline the voice input processing. Process 600 can then end at step 606.

[0055] FIG. 7 is a schematic view of an illustrative system for processing voice inputs based on the user's identity in accordance with one embodiment of the invention. System 700 can include some or all of the features of system 400 (FIG. 4), described above. System 700 can include voice input 710, which can include some or all of the features of voice input 410 (FIG. 4). Voice input 710 can be provided to voice input processing module 720 to identify one or more device operations to perform in response to the voice input. Voice input processing module 720 can include some or all of the features of voice input processing module 420 (FIG. 4). In some embodiments, voice input processing module 720 can include additional features not included in voice input processing module 420. For example, voice input processing module 720 can include one or more of software, firmware and hardware to perform user identification 722. In particular, processing module 720 can identify a user based on a user's inputs to a device or biometric information received form the user. For example, processing module 720 can detect a password or key known only to a particular user, detect an input for performing a device operation typically selected by a particular user, or receiving biometric data from an appropriate sensor.

[0056] Using user identification 722, processing module 720 can retrieve a particular subset 732 of words from library 730 for processing voice input 710 and identifying particular words or phrases of the voice input. Processing module 720 can provide user identification 722 to library 730 such that library 730 can retrieve a particular subset of library words associated with the identified user. Processing module 720 can then compare voice input 710 with library subset 732 to more efficiently identify specific words or phrases within the voice input (e.g., only comparing to the most relevant words or phrases, or most likely words or phrases to be used in the voice input). For example, voice input processing module 720 can identify one or more keywords specifying an instruction to the device, where the instruction can include one or more variables or values qualifying the instruction. The instruction (e.g., “play”), including the variables or values specifying how the instruction is to be executed (e.g., “Mika's latest album”) can be analyzed to identify one or more electronic device operations corresponding to the instruction.

[0057] Library 730 can include some or all of the features of library 422 (FIG. 4). For example, library 730 can be separate from voice input processing module 720, or instead or in addition embedded within voice input processing module 720. Library 730 can include any suitable words, including for example default words associated with the electronic device detecting the voice input, specific words derived from the user's interactions with the electronic device (e.g., with content transferred to the electronic device by the user), or other words or phrases. Subset 732 of library words can include any suitable subset of the library, including for example default words or user-specific words.

[0058] The particular words or phrases to place in subset 732 can be selected using any suitable approach. In some embodiments, processing module 720 can determine the user's interests 724 and select a particular subset of library words based on the user's interests. Alternatively, library 730 can receive users interests 724 from the processing module, or can retrieve the user's interests directly from the user or from an electronic device. Library 730 can then select the particular words or phrases to include in subset 732. Any suitable approach can be used to correlate a user's interests to words or phrases of a library. For example, words can be selected based on the types of applications or processes used by the user. As another example, words can be selected based on content consumed by the user (e.g., media items played back by the user). As still another example, words can be selected based on data used to perform one or more device operations (e.g., contact information of particular contacts to whom the user sends emails or messages).

[0059] Processing module 720 can identify the user's interests 724 using any suitable approach. In some embodiments, processing module 720 can receive user feedback 742 from electronic device 740. The user feedback can include any suitable type of feedback from which user interests 724 can be derived. For example, user feedback 742 can include playback information for media items (e.g., which media items are selected for playback, or skipped during playback), user interactions with the device such as user instructions relating to content accessed using the device (e.g., star rankings provided by the user for media items) or particular applications or operations that the user selects to execute (e.g., a particular game that the user plays), or any other feedback describing a user's interactions with the device. In some cases, user feedback 742 can be provided to library 730 instead of or in addition to processing module 720 for creating subset 732 in the library.

[0060] Voice input processing module 720 can provide an instruction derived from the identified words of voice input 710 to device 740. Device 740 can in turn identify one or more operations to perform in response to the received instruction, and execute the one or more operations. In some embodiments, processing module 720 can instead or in addition identify the one or more operations related to a derived instruction, and provide the operations direction to device 740 for execution. Device 740 can perform any suitable operation, including for example operations relating to one or more applications or processes within one or more applications, and can include a punctual, repeating, or lasting operation (e.g., monitor all incoming email for particular flagged messages). Device 740 can include any suitable device, and can include some or all of the features of electronic device 100 (FIG. 1). In some embodiments, device 740 can detect and provide voice input 710 to voice input processing module 720, which can also reside on device 740. In some embodiments, device 740 can instead or in addition be a distinct device that receives instructions to perform operations from a remote voice input processing module 720. Device 740 can receive instructions from processing module 720 and can provide user feedback to processing module 740 over path 754. Processing module 720 can compare voice inputs received over path 750 with library 730, and can assist in the selection of subset 732 via communications over path 752. Each of paths 750, 752 and 754 can be provided over any suitable communications network or protocol, including for example wired and wireless networks and protocols.

[0061] In some embodiments, the voice input can include a word defining an arbitrary or user-specific variable for a device operation. For example, the user can provide a voice input directing the device to play back and a media item that the user will find “good.” The processing module can use user's interests 724 to quantify abstract or qualifying terms and provide actual variables or arguments for the device operations. For example, the electronic device can select recently added or loaded media items, current hits or higher ranked media items, media items with higher play counts, or media items by a favorite artist or within a preferred genre.

[0062] The following flowcharts describe various processes performed in some embodiments of this invention. Although the descriptions for the following flowcharts will be provided in the context of an electronic device, it will be understood that a voice input processing module can perform some or all of the process steps. FIG. 8 is a flowchart of an illustrative process for performing a device operation in response to a voice input from an identified user in accordance with one embodiment of the invention.

[0063] Process 800 can begin at step 802. At step 804, the electronic device can determine whether a voice input was received. For example, the electronic device can determine whether an input interface detected an analog signal corresponding to a voice input. If no voice input is received, process 800 can move to step 806 and end.

[0064] If, at step 804, the electronic device instead determines that a voice input is received, process 800 can move to step 808. At step 808, the electronic device can determine whether the user providing the voice input was identified. For example, the electronic device can determine whether the user provided an input characteristic of the user (e.g., a user name and password, or using a particular application specific to a user). As another example, the electronic device can determine whether biometric information related to the user providing the input has been detected. The electronic device can compare the identification information with a library of authentication or identification information to identify the user. If the user is not identified, process 800 can move to step 810. At step 810, the electronic device can process the received voice input using a full library. For example, the electronic device can identify particular words or phrases of the voice input from an entire library of words used to process voice inputs. Process 800 can then move to step 810.

[0065] If, at step 808, the electronic device instead determines that the user was identified, process 800 can move to step 812. At step 812, the electronic device can identify a subset of a library used to process voice inputs. The identified subset can be associated with the identified user, such that words in the subset relate to interests of the user, or to words that the user is likely to use when providing voice inputs. For example, words in the identified subset can include metadata values that relate to content (e.g., media items or contacts) stored by the user on the device. At step 814, the electronic device can process the voice output using the identified subset of the library. For example, the electronic device can compare the received voice input with words of the subset, and identify specific words or phrases of the voice input. At step 816, the electronic device can identify an electronic device operation corresponding to the processed voice input. For example, the electronic device can identify one or more operations or processes to perform based on the voice instruction (e.g., generate a playlist based on a particular media item). At step 818, the electronic device can perform the identified device operation. Process 800 can then end at step 806.

[0066] FIG. 9 is a flowchart of an illustrative process for processing voice inputs based on a user's identity in accordance with one embodiment of the invention. Process 900 can start at step 902. At step 904, the electronic device can receive a voice input. For example, an input interface of the electronic device can receive a voice input using a microphone. At step 906, the electronic device can identify the user providing the voice input. For example, the electronic device can identify a user from an input characteristic of the user (e.g., a user name and password, or using a particular application specific to a user). As another example, the electronic device can identify a user from detected biometric information. At step 908, the electronic device can identify a subset of library words associated with the identified user. The subset can include words that relate to interests of the user, or words that the user is likely to use when providing voice inputs. For example, words in the identified subset can include metadata values that relate to content (e.g., media items or contacts) stored by the user on the device. At step 910, the electronic device can process the voice output using the identified subset of the library. For example, the electronic device can compare the received voice input with words of the subset, and identify specific words or phrases of the voice input. In some embodiments, the electronic device can identify an instruction to provide to a processor, or an operation for the device to perform from the processed voice input. Process 900 can end at step 912.

[0067] FIG. 10 is a flowchart of an illustrative process for defining a subset of library words related to a user in accordance with one embodiment of the invention. Process 1000 can begin at step 1002. At step 1004, the electronic device can retrieve or access a library of words for processing voice inputs received by the device. For example, the electronic device can access a library of words typically used for providing voice inputs to an electronic device. At step 1006, the electronic device can identify a user's interests. For example, the electronic device can review past user use of the device, user-selected content stored on the device, request preference information from the user, or review words identified from previous voice inputs. At step 1008, the electronic device can extract words that the user is likely to use to provide a voice input. For example, the electronic device can identify particular words that relate to the user's interests, or words that the user is likely to use based on the types of applications used by the user. At step 1010, the electronic device can define a subset of the library that includes the extracted words. For example, the subset of the library can include the intersection of the extracted words and of the library. As another example, the electronic device can identify words of the library that share a root or other common feature with the extracted words. Process 1000 can then end at step 1012.

[0068] Although many of the embodiments of the present invention are described herein with respect to personal computing devices, it should be understood that the present invention is not limited to personal computing applications, but is generally applicable to other applications.

[0069] Embodiments of the invention are preferably implemented by software, but can also be implemented in hardware or a combination of hardware and software. Embodiments of the invention can also be embodied as computer readable code on a computer readable medium. The computer readable medium is any data storage device that can store data which can thereafter be read by a computer system. Examples of the computer readable medium include read-only memory, random-access memory, CD-ROMs, DVDs, magnetic tape, and optical data storage devices. The computer readable medium can also be distributed over network-coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

[0070] Insubstantial changes from the claimed subject matter as viewed by a person with ordinary skill in the art, now known or later devised, are expressly contemplated as being equivalently within the scope of the claims. Therefore, obvious substitutions now or later known to one with ordinary skill in the art are defined to be within the scope of the defined elements.

[0071] The above described embodiments of the invention are presented for purposes of illustration and not of limitation.