JP2005209331A - Intelligent music track selection - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide intelligent music track selection. <P>SOLUTION: The user of a first electronic device is enabled to control the performance of multimedia works by the first electronic device, data permitting deduction of user's preference regarding the multimedia works therefrom is obtained by the first electronic device, and preference information is generated by another electronic device based on the obtained data, whereby the performance of the multimedia works is controlled based on the preference indicated by the user. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to selection of items based on user responses.

  This application is a continuation-in-part of US Patent Application No. 10/180900, titled “Selection of Items Based on User Reactions,” filed June 25, 2002, which is incorporated herein in its entirety.

  For example, for items that are tracks of music on a compact disc, some CD players allow users to indicate their preferences by programming a sequence to play the tracks. When a track ends, the CD player selects the next track based on the programmed sequence.

  In some music playback devices and music playback software, a user can indicate a type of music, eg, a jazz preference, and the device or software plays an item of the selected type of music.

  In one proposed system, the next track to be played is selected based on rating information supplied by the user. Some systems provide a thumbs up / thums down interface for users to indicate their preferences. In another system, the music item to be played is selected based on the preferences indicated by the listener at the thumb-up / thum-down interface. Some systems provide an automatic playlist function in which music items are selected for playback based on a user's answer to a question indicating a preference for the type of music. At least one online merchant infers the user's hobbies based on music purchased by the user.

  In other cases, the item selected is a function of the software application or a file used by the software application. Users express their preferences by using functions or files. For example, in Microsoft Word, recently used files are selected by the application to display in the recently used file list in the drop-down menu, and only recently used functions (eg outline functions) are available for user invocation. For the display in the drop-down menu is selected.

  In general, the present invention, in one aspect, allows a user of a first electronic device to control performance of a multimedia work (work) by the first electronic device; Acquiring (acquiring) data from which the reference can be inferred by the first electronic device, and generating preference information based on the acquired data by another electronic device, the preference Is characterized by a method comprising steps, including negative preferences.

  Embodiments of the invention include one or more of the following features. The preferences include only negative preferences. The data is acquired based on the user control of the performance of the work. The user is allowed to control the performance of the multimedia work with one or more other electronic devices, and the user's preferences for the multimedia work are controlled with the one or more other electronic devices. Data from which can be inferred is obtained, the preference includes a negative preference, and preference information is obtained from the first electronic device and the one or more other electronic devices. And the generated preference information is associated with the user. The first electronic device includes a portable device.

  A sequence of performances of the multimedia work is specified for the user based on the preference information. A multimedia work is selected for distribution to the first electronic device based on the preference information. A multimedia work is selected for distribution to at least one of the devices based on the preference information. The multimedia work is stored (stored) in the first electronic device. The multimedia work is stored on at least one of the electronic devices. A changing body of a multimedia work is stored at the first electronic device, and the body of the multimedia work is determined based on the preference information.

  The data acquired at the first electronic device includes data from which it can be inferred whether a user has not permitted the performance of the work to be completed. The data obtained at the one or more other electronic devices includes data from which it can be inferred whether a user has not allowed the performance of the work to be completed. The data acquired by the first electronic device includes a time when the performance of the work was last completed. The data acquired at the one or more other electronic devices includes a time when the performance of the work was last completed. If the performance of the work is not allowed to complete, it is inferred that the user dislikes the work.

  The multimedia work stored on the first electronic device includes a larger library subset of multimedia works available to the user. The acquired data is provided from the first electronic device to the other electronic device using an electronic communication channel or a portable storage device. The preferences include only negative preferences. The user is verified for identity and the preference information is associated with the user. Preference information derived from the user during the performance of the work on the first electronic device is received at an online sales site, and based on the preference information, a work recommended to the user for purchase is selected. Is done. At least one of the one or more other electronic devices includes a portable electronic device. The portable electronic device includes a compressed music player. At the online sales site, user preference information is received, and based on the preference information, a work proposed to the user for purchase is selected, and the work selected by the user in response to the proposal is Downloaded and fees are assessed for the performance of the downloaded work.

  In general, in another aspect, the invention features a method that includes maintaining each stored library of multimedia works at a library electronic device. A body of preference information indicating user preferences for the multimedia work is stored on the portable electronic device. A user is allowed to have the preference information stored on the portable electronic device control the performance of a multimedia work from any of the library electronic devices.

  In general, in another aspect, the invention provides user preference information for storing a library of multimedia works in a storage device and playing to a user at a location associated with the storage device. Selecting a portion of the multimedia work from the library based on a probability and assessing a fee for playing the selected multimedia work on a performance device.

  Embodiments of the invention include one or more of the following features. The selection of a portion of the multimedia work is based on a preference indicated by the user regarding the performance of the multimedia work. The fee is based on the number of works performed. The fee includes a one-time fee for all performances of the works in the library. The fee includes a fee for each use. The work is locked for access and the fee is assessed for unlocking access. The library is automatically updated on the storage device. The update includes adding and / or removing works in the library. Based on the information about the selected multimedia work, a royalty is paid to the rights holder of the work. The location includes a location associated with personal movement. The location includes one of a car, an airport, or a hotel.

  In general, in another aspect, the invention includes preloading a library of multimedia works into a vehicle storage device, wherein the multimedia works in the library are inaccessible except when unlocked. And a step of selectively unlocking a portion of the library based on preference information indicating preferences relating to multimedia works of the user of the vehicle.

Embodiments of the invention include one or more of the following features. At least a portion of the preference information is supplied by the user during use of the vehicle.
In general, in another aspect, the invention shows the steps of playing a first multimedia work for a user and preferences relating to the multimedia work played after the user resembles the first work. And a method comprising the steps of enabling:

  Embodiments of the invention include one or more of the following features. The user is allowed to show preferences for multimedia works similar to the first work to be played later.

  In general, in another aspect, the invention shows the steps of playing a first multimedia work for a user and preferences relating to multimedia work that the user played later does not resemble the first work. And a method comprising the steps of enabling:

  Embodiments of the present invention can include features that allow the user to indicate preferences for multimedia works similar to the first work that is played later.

  In general, in another aspect, the invention manages a body of preference information indicating user preferences for a multimedia work that is part of a first library of multimedia works; Applying the preference information to generate a performance sequence for a second library of separate multimedia works.

  Embodiments of the invention include one or more of the following features. The first library and the second library share at least some common works. The first library of multimedia works and the second library of multimedia works can be used at different locations. The preference information is stored on a portable medium and carried from one location to another for use with different libraries of multimedia works. The preference information is stored in the network.

  In general, in another aspect, the invention manages a first body of preference data or preference information indicative of a first user's preferences for a multimedia work, and a second user's preferences for the multimedia work. Managing a second body of preference data or preference information indicative of a reference; and deriving a third body of preference data or preference information from the first body and the second body. Features.

  Embodiments of the invention include one or more of the following features. The third body of preference information is used to generate a performance sequence for additional users. The derivation includes aggregating the first body and the second body of data or information and processing the aggregation to derive the third body of data or information. The first user and the second user include the same user.

  In general, in another aspect, the invention includes storing at a device a set of multimedia works including a sub-library of a larger library of available multimedia works, and at least some of the multimedia works at the device Playing data, receiving data indicating user preferences relating to the played multimedia works, and updating the multimedia works included in the stored set based on the user preferences And a method comprising the steps of:

  Embodiments of the invention include one or more of the following features. The device includes a handheld device or an in-vehicle device. The claimed method includes the steps of repeatedly receiving information indicative of user preferences and repeatedly updating the multimedia work based on the preferences.

  In general, in another aspect, the invention maintains a step of maintaining information indicative of a user's preferences regarding a played multimedia work, and a set of multimedia works based on the information and available multimedia works. Characterized in that it comprises the steps of stochastically generating and distributing the set of multimedia works to portable devices.

  Embodiments of the invention include one or more of the following features. The set of multimedia works is distributed in a compressed format. The compression format includes MP3. The distribution is performed by recording on a disc. The distribution includes downloading the multimedia work to a portable device. The download is performed in such a manner that the user does not notice that the download is being performed. The portable device includes a handheld device or an in-vehicle device. The download is performed wirelessly.

  In general, in another aspect, the invention maintains a body of preference information indicative of a user's preferences for a multimedia work, and at a performance device, receives preference information or work selection information from the user And selecting a multimedia work for performance based on the preference information or the work selection information or pre-stored preference information about the user's preferences regarding the multimedia work, comprising: A step in which a part of the selected work is not stored in the performance device and the unstored work can be used without necessarily indicating to the user that the work is not stored in the performance device Get it from the source It features a method comprising the step of loading the nest.

  In general, in another aspect, the present invention can maintain a multimedia work played on a portable device having limited storage and sometimes be preferred by a user on the portable device. The method includes a step of automatically replacing a work having a low probability with a work having a high possibility of being preferred.

  In general, in another aspect, the invention relates to a device that includes a storage that stores a set of multimedia works including a subset of a larger library of available multimedia works, and at least some of the multimedia works in the device Included in the stored set based on the elements configured to play the music, stored information indicating user preferences regarding the multimedia work being played, and the user preferences And a port for receiving an update of the multimedia work.

  In general, in another aspect, the invention relates to an online sales site for receiving preference information derived from a user during performance of a work on a local device, and for purchasing based on the preference information. Selecting a work proposed to the user.

  In general, in another aspect, the invention includes storing a library of multimedia works on a storage device, selecting a portion of the multimedia works from the library for performance to a user, Assessing a fee for playing the selected multimedia work on a device and preventing access by the user to data containing the played multimedia work.

Other aspects of the invention include combinations of other methods, devices that perform these methods, systems that include such devices, and means for performing the methods.
Other advantages and features will become apparent from the following description and from the claims.

As can be seen from the overview diagram 35, in the system 1500, the user 1502 is a multiplayer played for the user by local electronic devices 1504, 1506, 1508, and 1510, which are typically portable or mobile (which may not be mobile) Can have media work (works). Although only a few are shown, there can be thousands or millions of such devices operating in the system. One or more of various techniques may indicate that one or more users prefer a preference for a particular multimedia work or type of multimedia work, such as the Beatles album, at local device 1504. . Preferences instructions can be accumulated as preference data at each device. The accumulated preference data is generated by other devices (eg, other local devices or one or more host electronic devices 1512) to generate preference information that in some cases specifically associates that information with each user. , Updating, and storing (storing). Based on the stored preference information, the host or other device generates one or more playlists (a type of playback sequence) used by the local device to manage the playback of the multimedia work Can do. One or more hosts may also manage the downloading of multimedia works to local devices according to playlists and / or preference information.

  As a result, the user can enjoy a greatly enhanced experience, which is more consistent with preferences with respect to multimedia work. Various other features, described below, can also be used with the system 1500 shown.

  At least some of the devices 1504, 1506, 1508, and 1510 (eg, small handheld devices) may not have sufficient processing capabilities to run software that provides all of the multimedia preference system functionality. . In that case, a local device with limited functionality can perform only a few of the functions of the preference system, while other functions are performed centrally on the host. For example, a local device may capture some information, such as raw (raw) form of preference data, but another device uses that data to determine the piece to be played ( Other information (such as preference information) is derived by creating a playback sequence and downloading it to the local device and / or updating the downloaded library for use on the local device. In other examples, the local drive has the appropriate processing capabilities and coordinates with the host or other device, so that preference information is transferred to the device used by a particular user (or Update the library if the device has insufficient storage resources).

  The phrase “multimedia work” includes, when used alone, audio works (especially music), video works (especially movies), images, sounds, and works that can be experienced by human senses, especially visual or auditory (simple Included). When used with words such as audio or video elsewhere, a phrase multimedia work may mean a work that is not just an audio or video work.

  The phrase “local device” includes music players including MP3 players, mobile phones, personal digital assistants, radio, video, CDs, and other types of multimedia players in vehicles, ships, and airplanes, home-based music, Includes all types of multimedia players that serve video and other multimedia systems, and generally one or a few individual users.

  The word or phrase “preference data” or “preference information” includes an indication that the user likes or dislikes an individual multimedia work or type of multimedia work. For example, the preference data can be as simple as the user's choice to abort the performance of the work along the way. However, preference data can be complex and detailed. The preference information can be derived from the preference data and can be more abstract than the preference data. For example, if a user repeatedly stops playing a rock music piece, the raw data can capture individual events, but the preference information derived from the raw preference data can be Can hate you.

  Referring once again to FIG. 35, the handheld device 1504, in one embodiment, includes various so-called MP3 players that can store and play back audio works encoded in Apple's iPod or MP3 format under user control. can do.

  Device 1504 includes a microprocessor 1514 that executes software 1516 included in RAM 1518 (or firmware). Information about the piece being played or playable and other information related to device usage can be stored in RAM 1518 or mass storage 1524 and displayed to the user on display 1526. The storage device may be a disk drive, flash memory, or other form of storage (storage device). The user controls the action of the player using one or more control means 1528. The control can be a mechanical button or switch, a touch control accessible on the display, or by voice control or non-mechanical control of other types of devices. The storage device 1524 can be of a removable and reusable type from a device such as a removable disk drive or memory card, or it can be non-removable.

  Device 1504 may send multimedia work, preference data, preference information, and other information wirelessly between this device and other local devices or other devices including one or more hosts. Transceiver 1527 and associated antenna 1529 may also be included. A connector 1530 may be provided to allow similar transfers over a wired connection. Whether wireless transfer, wired transfer, or both are available, communication can occur over a dedicated channel or over one or more private or public networks 1532. Some versions of local devices may include hardware 1533 capable of playing audio and video, such as a DVD player.

  The storage 1524 can store various types of information. Multimedia works can be stored individually 1538, in a sub-library 1540, or in a library 1542. The stored work can be received on a removable medium or downloaded via a wireless or wired connection. The memorized work can be updated by adding additional works and removing the memorized work, and the updates can be made with the user's explicit preferences or captured in the preference data or preference information. It can be based at least in part on implicit preferences.

  Thus, swapping out a low preference work with a high preference work when the multimedia work stored on the local device has a lower relative preference score than other multimedia works available on the host The process can be carried out. Swapping increases the likelihood of satisfying users despite the limited storage of local devices. One way to control swapping is to remove the works with the lowest rating (up to a certain threshold, or up to a certain percentage of storage space) and replace them with the work with the highest rating from a larger library. Ratings were collected on other works using the score calculation engine by inserting metadata from unrated works into the metadata network (an example of a metadata network is described in detail below with respect to FIG. 12). The prediction can be made by determining the prediction score based on the information. In this process, an overlap between the unrated work and the rated work metadata is used to predict the rating of the unrated work.

  A playback history 1544 that includes the identifier of the played work, whether the playback is complete, and the time when the playback is complete can also be stored. The playback history may also indicate the identifier of the user 1502 who played the work, if the device maintains separate information for different identified users of the device. For example, Apple's iPod is configured to store the number of playbacks of a given work, the last playback time, and the rating (1-5 stars) of each work entered by the user.

  The storage can also include a predetermined playback sequence 1550, each of which identifies a sequence of works to be played. The playback sequence can be entered by the user or can be supplied from the host or another local device. If the local device on which the work is played has sufficient processing power, the playback sequence can be provided by the local device itself. In some cases, the playback sequence is determined based on the work available on the local device and historical preference data entered by the user and / or preference information associated with the user. A playback sequence can include a playlist that is an orderly simple list of works to be played, or a more complex list of items that control what is played, but not necessarily in a simple order.

  The storage can also include raw preference data 1552 associated with one or more users, or preference information derived from raw data as described above. The preference data can include a rating for each piece played on the device or available for play (eg, iPod star rating), or positive or negative preference for an individual piece or group of pieces Other types of instructions can be included.

  For example, the user can stop playing a piece in the middle, and from this the system can infer that the user has a negative preference for that piece (and possibly for a piece that is similar to it). The list of played works can be compared with the history of completed works to determine which works have been played but not completed. Conversely, if the user plays a piece completely, an inference can be drawn that the user has a positive preference for that piece (and possibly for a piece that is similar to it).

  Fetch and execute playback sequence 1554, decompression of compressed audio data 1556, playback history record 1558, preference data record 1560 (and, in some cases, preference) in RAM or other program storage in the local device. Derivation of preference information from reference data), interaction 1562 with host and other local devices, registration and login 1564 of different users, lock and unlock 1565 of played works, and libraries, sub-libraries, and Programs can be included that perform various functions, including individual work updates 1566.

  In addition to receiving and storing the work, the local device can export and import preference data and preference information for one or more given users on media 1570 (or electronically via a communication channel). Have the ability. In this manner, the user can carry (or electronically transmit) his / her preference information together for use with multiple local devices. For example, a user can bring his / her preference information into an airplane 1510, a car 1508, or a hotel room 1506 to control the playback sequence of local devices at these locations. Preference data and preference information can also be downloaded or distributed from the host to medium 1572. The preference data need not be stored or created on the same device on which the multimedia work is being played, but can be stored or created on separate devices.

  Works that are added to works stored on the local device can be downloaded electronically (eg, via either a wired or wireless connection) via network 1532 or distributed on media 1534 ( Distribution). The download can be done by a background process so that the user is unaware that the download is taking place.

  Host 1512 can be a workstation, server, or other electronic device and typically includes a microprocessor 1576, input / output devices 1578, RAM 1580, and storage 1582. The host may have a connector 1584 that downloads or exports work or preference information.

  Host storage can include a number of individual works 1586, as well as sub-libraries and libraries 1588 of individual works. Works can be obtained through a network connection of a host, such as a CD, DVD, or other available movable media, radio station, or television station, or an Internet website that provides downloadable media content It can be received from various sources 1590, such as other sources that can. The storage may also include preference information 1592 for each user and aggregated user information 1594 that may be useful for multiple users or groups of users. A playback sequence 1596 (including a playlist) generated by the host is also stored for download or distribution to the local device. The user information registry 1598 includes registration information and other information that allows the system to uniquely identify each user and interact with each user.

  Software executed on the host may include playlist generation software that generates a playback sequence including a playlist from the preference information and a list of available works. Library generation software 1702 manages the receipt and creation of works libraries and sub-libraries from which works can be downloaded or distributed to local devices. The update software 1704 manages the update process. Download software 1706 manages download or distribution processing. The preference generation software 1708 (described in detail below) uses preference data from one or more users to infer preference information for use in creating storage and playlists, expand. The fee assessment software 1710 manages the process of charging the user for the work played on the local device. Conversely, the rights management software 1712 manages the processing of payment of remuneration to the rights owner regarding the use of the work. Compression software 1714 manages the compression of works for distribution and download.

  The choice of whether to convert the preference data to preference information either entirely on the host, completely on the local device, or a combination of the two is a relative choice between the two devices. Depends at least in part on ability.

  Preference data entered at a local device (eg, portable music player) can also be used in connection with recommendations for works purchased by a user (additional information regarding purchase recommendations will be provided later). Purchase recommendations can relate to the entire CD, the entire DVD, or other compilations of the work, or can relate to individual tracks. Once preference data from the local device has been converted to preference information, it can be used to generate purchase recommendations. In one embodiment, preference data accumulated by a handheld music player such as Apple's iPod (or preference information derived from that data) is uploaded to a music retail (sales) website such as Apple's iTunes. can do. On the website, preference data can be converted into preference information. The converted or uploaded preference information is used by music sales site users to recommend edits or individual works (tracks) that the user may wish to purchase for download to his music player. Can be used to control the features. In another embodiment, preference data is converted to preference information prior to transfer to a music sales website. The conversion from preference data to preference information can be performed on the local device or host, or can be performed partially on the local device and partially on the host.

  In addition, the sales website may provide an automated download (directly or through another server or personal computer accessible to the user) to the user's local device of a preview portion of the work that the user has determined is of interest. Indirectly). Such a preview can also be part of a playback sequence that is downloaded to the local device. When a local device (e.g., a player) plays one of its parts, the performance is described in "The complete track is available for purchase from iTunes." Can be purchased from iTunes. If you want to buy this track, press the enter button) and you can continue with an audible or displayed message. The player records the purchase request and, if the player is connected to the website (or a network that can connect to the website), immediately forwards the purchase request to the sales website, or The request can be stored in memory to respond to a request when connected to a network or website.

  The interaction with the sales website can reach the user's local device via the computer via the host shown in FIG. In that example, in other embodiments of the local device of FIG. 35, the computer also stores downloaded work, playback sequences, and preference information and transfers them to the local device at a convenient time and place. By acting as a proxy for the local device, it can act as an intermediary between the host and the local device. The preference data and preference information uploaded to the host can also be temporarily stored in the computer. The host can be configured to interact with the sales website by supplying the user's preference information to the website automatically or when requested or approved by the user. The host can also act as an intermediary on behalf of the user to buy works that are downloaded to the local device.

  In another embodiment, the interaction between the sales site and the host or local device (including downloading of purchased work) occurs directly when the local device and the sales site or host communicate over the network. Can do. In some embodiments, the purchase request can be stored at the local device and later uploaded to the host for processing and requesting.

  If the local device is a system that is rented, temporarily used, or owned by a party other than the user, the actual works played and the edits to which these works belong are also proposed for purchase. include.

  Numerous details of host and local device embodiments and the form of deriving preference information from user input and generating playlists are included in the following description (and are hereby incorporated by reference in their entirety). Cited from US patent application Ser. No. 10/180900, filed June 25, 2002, many of which are listed below).

Among the many examples in which the system 1500 can be used to improve the local device user experience are:
A user of the local device can trigger playback of a sequence of songs. The playback sequence is predefined and stored on the device. For each song, the user can play the song to the end (in this case the fact and completion time are remembered), or the song can be stopped before it ends and proceed to the next song . In that case, nothing is recorded in the playback history (at least in the case of iPod). The local device or host can infer that the user does not like the song from the user's decision not to finish playing the song. It is possible to modify the playback sequence or configure the playback sequence later to provide the user with songs that are likely to be enjoyed by the user. In some cases, the only information regarding the user's preferences may be negative information inferred from the inability to complete playback of the song. In other cases, the user can provide positive and negative explicit feedback, eg, by a rating system.

  In some cases, a given user can show his preferences via multiple local devices, such as portable MP3 players, home audio systems, and rental cars. By aggregating, analyzing and deriving preference information from the preference data, the system can control the playlist more effectively for its user at each of the local devices. In addition to improving the playlist, the choice of which works to download or distribute to each of the local devices, initially or with updates, can be important. The augmentation of works stored on a given local device may be limited by the storage space, so the downloaded or distributed work may be only a subset of the library stored on the host.

  When the system is configured to charge the user for a performance performed, the fee is based on a single payment for each performance, for each use, or for all performances of a work in a given library. Can be assessed. Access to the work can be locked electronically and can only be unlocked when the performance of the work is paid for at an appropriate fee. In this system, for example, it is possible to download a huge library of works to the storage included in the rental car. Only the paid library or work is unlocked and made available for playing in the car. Such a library can include, for example, 10,000 or more works. In a system configured to charge a user for a played work, user access to actual data including the work can be prohibited for accesses other than the performance. Permission to modify a work that exists on the device's storage system is limited to the work's owner if the modification includes deleting the work, adding a work, copying, or operations on data or data files, Limited to those who are explicitly granted access to the data by the owner of the work.

  Apply user preference information derived from the use of one or more local devices (which may or may not be owned by the user) on other local devices (including those not owned by the user) Thus, it is possible to present to the user a work that the user is likely to like without having knowledge of the work stored in another local device or a host serving the local device.

  In another example of the use of a local device, a user may prefer to play another piece that resembles (or does not resemble) another piece, such as a piece that is currently being played or has just been completed. The device can be configured so that it can be explicitly shown. The user can enter the information via the user interface.

  Preference information developed during playback of a work from a library of works can be applied to other libraries, and two libraries can have a work in common.

  Multiple different users can generate a body of preference information or data, and the system aggregates these to produce a playback sequence and download of the work for use by each user or other users A shared body of preference information that can be used to generate can be generated. For example, a set of alumni can agree to pool preferences in order to generate a playback sequence that represents the joint preferences. Some aggregated information can be used by users who were not members of the original group. In the aggregated preference information, preference information of a certain user can be generated for a library of works different from the preference information of the second user. One form of aggregating preference information is, for each user, a list of works played, actions performed on them (played, skipped, rated), or basic feedback behavior To maintain a list (track X: implicit affirmation; track Y: implicit negation;...). In order to merge the data, both sets of feedback data are performed through the same metadata network described below.

  If a personal device has limited storage, a subset of the work stored on the device (from a larger library) is adjusted by adding and removing works over time to reflect the user's preferences can do. Thus, the device maintains a changing subset of the larger library that is updated repeatedly. The work stored on the local device at a given time has, on average, an ever-increasing potential for user preference.

  Updating the work sub-library and playback sequence from the host to the local device can be done in the background without the user explicitly knowing that the transfer is taking place.

  When a user uses a local device, the user logs on to the device in such a way that the device or host can identify the user and sets one or more sets of preference information for the user's session with the local device. Can be used. A user may be allowed to select a set of preference information from a list.

  For example, an embodiment of a system used with an iPod may be based on publicly available information about a database maintained by the iPod. See, for example, http: // sourceforge., Which is incorporated herein by reference. net / docman / display_doc. php? docid = 11191 & group_id = 52976, http: // sourceforge. net / docman / display_doc. php? docid = 111212 & group_id = 52976, http: // ipod-on-linux. sourceforge. net / docs. php, http: // neuron. com / ~ jason / db. py via, and http: // neuron. com / ~ jason / ipod_archive. See html.

  This publicly available information indicates that the iPod database is split into two files, called “itunesdb” and “play counts”. “Itunesdb” includes a file name, track number, track title, album title, artist name, composer name, genre, and a list of playlists stored in the iPod (among other information). A list of stored tracks is included.

  The “play counts” includes one entry for each track, and this entry stores the last reproduction time, the number of reproductions, and the track rating (based on the 5-star system). This file is created by iPod and is updated whenever a track is played or the user changes the rating.

A simple program that uses the information in the database works in the following way:
Read “itunesdb” and get a list of all tracks along with their filename, album title, artist name, and genre.

Based on these, a metadata network is created, starting with the genre, followed by the artist, album, and file name.
Create a list of all playlists in the ipod, including rebuilding the list of each album based on the metadata in itunesdb.

  Read “play counts” and sort in order of last playback time. For each sequential pair of tracks in the last playlist, look for the smallest gap (in terms of the number of tracks in the middle) between the tracks in the playlist. Each pair of tracks is given positive implicit feedback and all of the tracks in the gap are given negative implicit feedback. Once all tracks have been rated, positive / negative implicit feedback is given as appropriate. All of this feedback is stored in a data file and accumulated over time.

Process all previously stored feedback over the network.
Run a playlist generation algorithm in batch mode (see discussion below) to generate a new playlist, ie select a number of tracks sequentially, as if each track is played without feedback Insert the playlist into “itunesdb”.

Additional information useful for understanding how to implement the techniques described above is provided below.
As can be seen from FIG. 33, in one embodiment of the system 10 for selecting items based on user preferences, the item is a piece of music (track) that is part of the supply 12 of available tracks.

  The track is supplied from a track source 14, which may be an audio tape, a compact disc, a mini-disc, a digital versatile disc, or a broadcast medium, wireless connection, satellite radio, the Internet or other computer network. Media such as other streaming media, or other sources, including other sources. Thus, the term track is used as if it only appears as a track of tangible media, but whatever the term track is created or stored or distributed to the user, all items It is intended to include.

  Also, much of this discussion uses music examples, but the items for which preferences are expressed include, among other things, all types of items, including audio, video, and multimedia items, database records, and the World Wide Web. It can be an item that can be obtained by searching on or a non-audio visual item that includes features of a software program. Similarly, all types of track sources 14 are possible, including, for example, commercially available sources and sources created by users. The terms CD and compact disc are used to include sources of all types of tracks. In some cases, when a track represents a characteristic of a software program, the source of the track becomes the system itself (since the software is the “source” of its own characteristic). Instead, in this example, the CD can represent, for example, a logical subset of software features, and all text formatting commands constitute a single CD, with each command on a single track. Become.

  The available truck supply can be stored in mass storage, so that the supply can be used locally as needed. Storage devices can include hard disks (including portable hard drives), CD-RW drives, tape recorders, video recorders, memory devices, or other storage devices that can hold multiple items for performance to the user. . For simplicity of explanation, all storage devices that provide read access (with or without write access) are referred to herein instead as hard disks or storage devices. In some cases, it may not be necessary to store available tracks locally. Instead, it may be possible to make available tracks available on-the-fly from remote sources via wired or wireless communication channels. When storage is used, the system can provide a mechanism for capturing tracks, converting them to the appropriate format, generating associated identification information, and recording them on a storage medium.

  In the system shown in FIG. 33, a track is played for the user by the track player 16. The track to be played at a given time is identified by the next track identifier 18. A track player can include any device that can play a track for a user in response to a signal identifying the track to be played. Therefore, for example, a CD player, a CD changer, a tape player, a wireless music playback device, an MP3 player, a video tape player, a DVD player, a web browser, a display device, a mobile information terminal, a mobile phone, a fixed phone, or a computer is used as a player. Can be included. The nature of the track player depends on the nature of the track. For example, if the track represents a characteristic of the software program, the track player becomes the software program itself. In some embodiments, multiple track players can be controlled by a track selector.

  When the player receives the next track identifier from the track selector, the track player removes the identified track from the supply and plays for the user 20. In the case of audio, video, or other multimedia tracks, the playback of the tracks is usually done sequentially. For example, when tracks include records in a database or resources available via the Internet, tracks can be displayed in groups simultaneously based on a set of next track selectors. Accordingly, this specification broadly means that the term playback or performance is used to display, perform, or otherwise distribute items to the user in a manner appropriate to the track and the user.

  The track selector 22 determines which track is to be reproduced next. The track selector can take a variety of forms. In some cases, the track selector is a general purpose computer that includes a microprocessor, memory, appropriate buses and caches, mass storage, operating system software, application software, and input / output drivers for the associated input and output devices. It can be. In other cases, the track selector may be a set of custom builds of circuits designed for the specific purposes described herein. The track selector can be part of an existing system in a vehicle or building or in a computer or handheld device. The entire set of elements shown in FIG. 33 can be implemented in a handheld device such as a personal digital assistant, mobile phone, or music player, or in a system located in a vehicle or building.

  The track selector can include logic 24 in the form of software, firmware, or hardware and storage 26 in the form of memory or mass storage. Track selection is performed by the track selection module 28. Details of the track selection process will be described later, but in general, the track selection is based on information supplied by the user via the user interface 30 and information about the tracks held in the storage 26. In some cases, the information supplied by the user is an implicit or explicit feedback that indicates the user's reaction or preferences regarding the track and the selection of the track to be played in dependence on that feedback. The reaction information can be raw information entered by the user, or information derived from the raw information (eg, variables), but used for purposes including selecting tracks to be stored and played in the future Is done. In other cases, track selection is not based on past user preferences or reaction information, but only on a more common set of commands supplied by the user via the input side 32 of the user interface. Ordinary commands can be provided, for example, via a user interface aspect that allows the user to specify that the system wants the system to play all tracks on the selected CD in track order. .

  The user interface includes both an input side 32 and an output side 34. The input side can include buttons, knobs, touch panels, voice recognition software, and sensors that can implicitly determine what the user wants the system to do or the preference information the user is trying to supply. The output side can include audible signals, displays, lights, speech synthesis, or other devices that provide information to the user regarding the state of operation of the system.

  On the input side, explicit and implicit mechanisms can be included where the user indicates preferences about the track and otherwise provides feedback or reaction. Explicit mechanisms can include buttons or other devices that allow the user to indicate directly whether he likes or dislikes the track, eg, the track being played. An implicit mechanism can also be used, for example, when the user presses the transport button to skip to the next track, this action can be interpreted as an implicit indication that the user does not like the playing track. it can.

  The track selector storage holds information that allows the system to select a track for playback. Track scoring information 36 may be provided for all tracks in the truck supply. The track scoring information for all tracks in the track supply is updated each time the track is played or the user takes actions via the user interface that reflect user feedback and preferences. Tracks are selected based on relative scores, as will be described later.

  The storage 26 also holds a playback history 38 that records a sequence of previously played track identifiers and metadata 40 that holds information about the track (eg, title, performer, genre). Instead, the metadata can be located in a read-only repository separate from the storage, such as a CD-ROM, or can be accessed via an over-the-air mechanism from an external central data source. Other information is also stored in the storage, which will be described in detail below.

  In addition to the track selection module 28, the logic 24 includes a module 50 that controls interaction with the user interface, and a module 52 that controls storage of the track to a supply of available tracks. The metadata maintenance module 56 retrieves and maintains metadata, which is supplied from a metadata source 60 and in some cases directly by the user, via user interface input, or storage Can be supplied via a personal computer interface. The playback history maintenance module 58 maintains the playback history 38. The track scoring module 54 can maintain data regarding user responses to previously played tracks and can assign scores to the tracks for use by the track selection module 28. Other logic modules can be included as will be apparent from the description below.

  Item selection system 10 provides the user with simple and intuitive access to items such as music tracks that satisfy the user's preferences without the need for prior configuration. The system automatically selects items or tracks from the supply, taking into account past and current user behavior and internal knowledge of the item contents. This system is arranged to simplify the provision of feedback from the user. With minimal intervention from the user, the system selects items that the user is interested in.

  In a simple example of the operation of this system, when listening to a track, the user can press a preference button that indicates that the track is preferred or another preference button that indicates that the track is disliked. . In either case, information about the user's preferences is accumulated and stored as part of the track's score. At the same time, the preference information is used to change the score of other tracks, for example another track of the same CD or another track of another CD by the same artist. When the user indicates he dislikes a track, the system jumps to another track. The user can again indicate his preference by pressing one of the preference buttons, and the information is again used to re-score the track. The implicit preference information is also used when the user plays the track completely or when the user presses the transport button to force the system to jump to the next track.

  When the system is operating in a mode that automatically selects a track, each track in the supply of available tracks has a non-zero probability that it will be played next according to an algorithm that considers the track's score. . Thus, the system does this quickly and effectively adapting to user preferences and requiring only a minimal amount of interaction by the user indicating the user's preferences. Details regarding how the user represents preferences, how the system selects tracks for playback, and how track scores are maintained will be provided later.

Among the advantages of the benefit selection system are one or more of the following.
(1) Simplicity. The system provides easy access to a collection of tracks. The system allows the user to simply press the “Power” button and begin listening to the tracks that the user enjoys.
(2) Control. The system provides this easy access without the user losing control over what is played. By representing a simple response to the currently playing track, the user can influence what the system chooses for the next playback. In the music field, CD players offer a high degree of control but require significant user intervention in the music selection process, which can reduce the overall simplicity (plainness). The radio requires minimal user intervention and allows a high degree of simplicity but allows little control. See FIG. When applied to music tracks, this system is intended to fill the gap.
(3) Access. The system provides access to the entire stored collection of tracks without necessarily requiring that the original media be usable.
(4) No setup. The system provides this control without requiring the user to determine what to play in advance.
(5) Safety. The system allows the user to enjoy the entire collection while driving a car safely or doing other tasks that require special attention.

In the context of platform music items, the selection system can be implemented with a hardware system of the form shown in FIG. System 708 includes computer 704, storage device 705, amplifier 706, and speaker 707. External components can include a CD drive 701, user input 702, and a display 703 for the user.

Three possible platforms for integration of the selection system are shown in FIGS. 1, 2, and 3, although various platforms can be used.
As can be seen from FIG. 1, the system 103 can be incorporated into a home entertainment system such as the Bose Lifestyle 35 (106) by adding a hard drive 101 for storing music for later playback. The system determines the sequence of tracks to be played and supplies the sequence from the hard drive to the entertainment system. User preference information and feedback are received from the user at the integrated remote control 107 and sent back to the system 103 for use in determining which track will be played in the future. Entertainment system 106 has the ability to play music from AM and FM radio 104, compact disc or DVD 105, or its hard drive 101, is multi-zone capable, and plays different items of music in different areas of the home. Can do. Remote control 107 interfaces with both core unit 108 and item selection system 103.

  In another embodiment shown in FIG. 2, the item selection system 202 is hosted in a smaller unit, such as the Bose Acoustic Wave Music System 205, also by adding a hard drive or other random access storage medium 201. can do. In such a configuration, the system can play music from the AM and FM radio 203, the compact disc 204, and the internal hard drive 201. Remote control 206 interfaces with both core unit 207 and item selection system 202.

  In the third embodiment shown in FIG. 3, the item selection system 305 can be integrated into the vehicle's music network. In this embodiment, the system includes a standard automotive entertainment system 309 (including, at a minimum, an audio source such as a compact disc player 302 and speakers 303), a hard drive or other random access storage medium 304, a computing engine 307. , And additional user interface controls 306 are augmented and linked to the audiovisual network of the car. In this configuration, the system plays music from its internal hard drive 304, which can record music from AM and FM radio 301, CD or DVD 302, and potentially satellite radio 308.

  As noted above, this system can also be implemented on other platforms such as video jukebox systems, database engines, or software programs (the selected item can be a feature or menu item). .

As stated before the mode of operation , the system can play tracks from a variety of sources, for example:
(1) CD / DVD. The system plays the currently loaded CD / DVD disc. In this mode, for the currently loaded CD / DVD disc, the user can select standard options available on the compact disc player, eg play, pause, next track, previous track, etc. . The operation in the CD / DVD mode is continued until a new mode is selected.
(2) AM / FM / satellite radio. The system operates as standard AM / FM radio or satellite radio.
(3) Cassette tape. The system operates as a cassette player.
(4) Stereo track. The system will play tracks from its internal hard drive.
(5) Internet. The system plays a track retrieved from another computer located on the Internet.
(6) Inside. For example, external data may not be required if the track is a feature of computer software.

When the system plays a track from its hard disk or when no external data is needed, the system can operate in one of the following modes:
(1) Automatic selection mode. As can be seen from FIG. 6, in this mode, the system plays a selection taken from all tracks available to the system. The user provides explicit or implicit feedback regarding the track being played. The feedback can represent user preferences. The system selects tracks for the user based on user feedback (emphasizing more recent feedback) and the history of recently played tracks. The selection system can also take into account feedback stored in current so-called user stations (see “Stations and Presets” below). In some embodiments, to increase simplicity, the auto-select mode is the default mode and the system returns to auto-select mode when other modes have completed playback.

  FIG. 6 illustrates the process by which the selection system functions when in automatic selection mode. Track selection is accomplished by an automatic selection module (see “Module” below). The auto-select module is initially in an idle state 2201 (note that other modules may not be idle, specifically the system may be playing a track). When the module receives a notification 2204 that it is time to select the next track (a signal can be provided by the player indicating when the track is complete or soon), in the "Module" section As will be described, the module switches to a “track selection” state 2202. The module selects the next track to play as described in a later section. If a further track selection request is received while the module is in the track selection state, the module adds the request to the track selection queue. If the module completes track selection and there are no items in the track selection queue or feedback queue, the system returns to the idle state (2205). If there are more track selection requests, the system returns to the track selection state (2206).

  If the module is idle when it receives a feedback event (ie, a message from the user interface module that there has been an implicit or explicit indication from the user regarding the currently playing track) The system switches to the “update feedback structure” state 2203 (2209). If the module receives a message while it is in the track selection state or the feedback structure update state, the module places the event in the feedback queue. When the module completes the track selection process, if the feedback queue is not empty and there are no further track selection events, the module switches to a “feedback structure update” state (2207). In that state, the module updates the stored track graph as described below. When finished, the module returns to the idle state if the feedback queue and track selection queue are empty (2210). If the feedback queue is not empty, the module returns to the feedback structure update state (2211). If the track selection queue is not empty, the module returns to the track selection state (2208).

  In other embodiments, the user is allowed to specify that the automatic selection mode must operate at the CD level or album level rather than the track level. In this configuration, in the automatic selection mode, the CD that is expected to be of interest to the user is selected and the CD is played from the beginning to the end.

  (2) Track mode. In this mode, the system plays only pre-specified tracks, for example, only the tracks listed in a given preset defined by the user, or all tracks of a particular CD or album. This mode can have multiple playback options including (a) normal, (b) shuffle, and (c) repeat.

  (A) The “normal” option causes the system to play all tracks in the order programmed by the user. At the end of the last track, the system can return to automatic selection mode.

  (B) The “shuffle” option causes the system to play tracks specified by the user in a random order. When all tracks have been played, the system can return to automatic selection mode.

  (C) The “repeat” option causes the system to remain in track mode until the user instructs the system to change. This mode can be used with the “Normal” or “Shuffle” option to determine in what order the tracks are played before being repeated.

  (3) Bin mode. In this mode, each preset (represented by a user interface object such as a keypad button on the control panel) includes one or more CDs. The system plays all the CD tracks that the user has added to the current preset. If the CD is only partially stored, the system will only play the stored track. In a system that implements the bin mode, a CD can be added to and deleted from the bin preset via the setting menu. In some embodiments, when the system is playing a stored track in some mode, the user presses a preset key to add a CD from which the current track is recorded to a preset. Can be lowered and held. The order in which the tracks are played is specified by the same playback options described in “Tracks”. At the end of playback with the “normal” or “shuffle” option, the system can return to the automatic selection mode.

  (4) Virtual CD changer mode. This mode is similar to the bin mode, except that each preset represents only a single CD. Accordingly, the user can select a CD to be played by pressing a button on the keypad 1805. Instead, the user can enter a two-digit or three-digit code to select a particular CD. The user can also select one of the playback options described in “Track”. This option applies only to the current CD, but with an additional option, these can be applied to all tracks on any CD in any of the presets. At the end of playback, the system can return to automatic selection mode.

  (5) Search mode (see FIG. 9). In search mode, the user can scroll through the displayed list of available content sorted by disc number, artist, or other available information. The user can select the entire disc, artist, genre, or any combination for playback. Instead, the user can enter a code representing the CD recorded on the hard disk of the system and play the track of that CD.

  The playback options described for the track mode also apply to the search mode. At the end of playback with the “normal” or “shuffle” option, the system returns to the automatic selection mode. The playback is actually performed in the track mode, and in the search mode, a special playlist for playback in the track mode is simply set up.

  In one example, the system maintains only one playlist and the user activates this playlist by pressing 0 on the keypad. In this example, switching between some modes is performed according to the state diagram of FIG. The system starts in automatic selection mode 2001 by default. The user can change the user station by pressing the corresponding button on the remote controller keypad 1805 (see FIG. 18) (see “User Stations and Presets” below) (2005). ). When the track is complete (2006), the system remains in auto-select mode and selects a new track. When the user presses changer 1806, the system switches to changer mode 2003 (2007). In changer mode, if the user selects a CD by pressing a keypad button or entering a code directly, the system will remain in changer mode 2003 (2012) and the CD or keypad with the code specified. Playback of the track from the CD assigned to the button (referred to herein as a slot) is started. At the end of the track, the system plays the next track (2011). When the CD is completed, the system starts playing the CD specified by the next slot (2013). That is, when the user presses the keypad button 2, the system plays the CD in slot 2 and then plays the CD in slot 3. If the user presses “Smart play”, the system switches to automatic selection mode 2008. When the user presses Whole Album (all albums) from changer mode, the system stays in changer mode (2023) but starts the current album from the beginning.

  When the user presses search 1812 (and selects from the search screen) and presses play list 1807, the system returns from automatic selection mode 2001 or changer mode 2003 to track mode 2002 (2009, 2010). . From the automatic selection mode, the Whole Album button causes the system to switch to the track mode (2009). When the search button is pressed, the system sets up a playlist that represents the search results. When the playlist is pressed, the system plays the playlist track. When Whole Album is pressed, the system sets up a playlist that represents each of the tracks of the CD that are currently being played. In the track mode, the playlist is played as described above. Playback continues as long as there are tracks to play (2016). If the user presses “Smart Play” or if the system has completed all the tracks in the playlist and the previous state was auto play mode, the system switches to auto select mode (2014). . If the user presses "Changer" or if the system has completed all the tracks in the playlist and the previous state was changer mode, the system switches to changer mode (2015). .

  If the user presses the another source button from any of the three states, the system switches to the new source 2004 after saving the state (2018, 2020, 2021). From other sources, the user can have the currently playing track recorded on the hard disk of the system. When the storage is complete, the system returns to the previous state (2017, 2019, 2022). If the user presses “Smart play” from another source mode, the system switches to the most recently used user station (2019). If the user presses “Changer”, the system switches to changer mode (2022). If the user presses the playlist, the system returns to track mode.

  In this embodiment, the current track is added to the playlist by holding down 0 while the track is being played in changer mode, track mode, or auto-select mode. The user can remove a track from the playlist by pressing the remove button or selecting an option from the settings menu.

  In other embodiments, multiple playlists are allowed, and in some embodiments, the “bin” described above, possibly by modifying the track mode to play the entire CD rather than individual tracks of the CD. Mode is also implemented. Some embodiments do not include some of the modes listed above. Other embodiments can include additional or replacement modes. For example, in some embodiments, the “all items” button is similar to the “where album” button except that it plays all tracks played by the same artist as the currently playing track. By this artist (all items by this artist) "button can be included. Other subsets of search functionality are possible, such as “all items in this genre” (all items in this genre).

Dynamic complexity adjustment selection system reduces the number of features available to the user, for example allowing only complex user operation modes, allowing for full control of all functions by the user, as well as simple user control, for example. Both automatic modes can be provided. In situations where the user's attention is primarily focused on another task, some features, such as complex aspects of the user interface, are disabled. During that time, the system provides a simple but functional interface that can satisfy the user's preferences without distracting the user's attention from key tasks.

  One example of this feature can be used in a selection system installed in an automobile (see FIG. 17). The system can be configured such that the search mode is not available to the user while the car is running. This mode can be limited to allow only when the vehicle gear is parked or the parking brake is engaged. Other modes can be used while in other gears.

Personal Computer Connection Some embodiments of this system may allow an optional interface with a personal computer. In these embodiments, the user can remove the storage device from the system and attach it directly to the computer. The user can save a list of all stored tracks with a code representing each CD. The user can print out the code associated with each CD for quick access after using the direct access function.

  While the hard disk is connected to the computer, the user can use the computer hardware to store additional tracks on the hard disk. This is advantageous, for example, when a computer CD player can read and store CD data more quickly than a system player.

  The computer can also provide the ability to create playlists and look up stored track metadata from metadata databases not available in the item selection system. Finally, the user can save the stored track collection on his computer for backup.

User interface The user interacts with the selection system through an interface mechanism that depends on the platform on which the system is installed. For example, as can be seen from FIG. 18, the user can use remote control. Alternatively, as shown in FIG. 23, the system can be controlled by a control panel when it is incorporated into a home entertainment system, for example. When installed in an automobile, the system has controls located on the steering wheel as shown in FIG. 24 and the remaining controls located on the dashboard as shown in FIG. be able to. When the system controls the selection of features of the computer software, the user interface is provided by the software itself. Some of the user interface elements described below may not be applicable to non-music tracks.

  Although the remote control of FIG. 18 shows most of the functionality available to the user, the control itself can be placed on the console, as a selection from an on-screen menu, or elsewhere. . Power can be controlled using a standard on / off button 1816. The user can mute the sound using the mute button 1818. The user can select the source of the track using a set of buttons 1826. Instead, a single button can be provided to cycle through the sources available to the system. When the system is in a source mode other than a stored track, the user can press and hold the “Stored” button 1817 to instruct the system to store the current track on the hard disk. it can. Instead, as in FIG. 25, the system can include a separate button 2503 that allows the user to save the current track. The system can also include a button 2502 that allows the user to delete the current track when the system is playing a stored track. Instead, this button can provide the functionality to delete all tracks from the current CD or album.

  A user can access advanced settings of the system using a settings button 1801. When in DVD mode, special features and information encoded in the DVD can be accessed using the DVD menu button 1804. Any of these menus can be exited using the exit button 1803. Navigation in these menus is accomplished using the arrow keys 1825. The enter key 1802 is used to accept choices in these menus.

  The user can control the volume using the volume key 1826. The button 1827 can be used to change the current user station (see “User Stations and Presets”) when operating in auto select mode, and is an input mode that supports track changes. Sometimes you can change the currently playing track. The current user station can also be changed by selecting a new station directly from the keypad 1805.

  From all modes except the playlist mode, the user can add a track to the playlist by pressing and holding 0 while the track is being played from the system hard disk. The user can enter playlist mode by pressing 0 once on keypad 1807. The user can return to the previous playback mode by pressing a button 1819 designated for this purpose.

  Control over the currently playing track can be exercised using standard transport controls. The user starts playback using the play button 1820, pauses playback using the pause button 1822, and stops playback using the stop button 1821. Can do. Further, the user can skip to the next track using next track button 1823 and return to the previous track using button 1824.

  When playing a hard disk track, the playback mode can be adjusted using buttons 1811, 1806, 1810 and 1813. The whole album button 1811 causes the system to play all tracks of the album from which the current track was recorded. The changer button 1806 causes the system to switch to virtual CD changer mode (see “Modes of Operation”). A smart play button 1810 causes the system to switch to automatic selection mode. The bins button 1813 causes the system to switch to bin mode. The user can enter the search mode by pressing the search button 1812.

  The system includes a feedback mechanism that allows the user to indicate a reaction (preference) to a track, eg, the currently playing track. This can be accomplished by providing simple + rating buttons 1815 and -rating buttons 1814. In another embodiment, as shown in FIG. 23, a + button 2302 is placed over the − button 2303 and the two buttons scroll the preference value between the maximum and minimum values. Can work as. In other embodiments, as shown in FIGS. 24 and 32, the + button 2401 may be larger than and above the − button 2402. In some embodiments, these buttons serve two purposes: a rating button when the system is in smart playback mode and a scroll button when the system is in other modes of playback. Other indicators can also be used. For example, instead of + and-, the system can be provided with a thumb up indicator and a thumb down indicator. The buttons can also be color coded (eg, red button for negative response, green button for positive response).

  In addition to the feedback button, some embodiments of the system can have an undo button. This button completely undoes the effect of the previously activated feedback button on the stored track graph (see “Metadata”). It is also possible to allow the user to go back to the previous track and reduce the effect of the accidentally entered preference display by showing the opposite feedback to that given in the previous playback of that track.

  Various mechanisms can be provided in the interface to allow the user to show positive or negative responses to the track. More than two buttons can be provided to allow explicit indication of preferences from more than two options. Conversely, a single feedback button, either a single positive feedback button or a single negative feedback button, or a single button that can be pressed for a certain amount of time to indicate the user's preference level Can be used. Buttons that are not simple but intended to provide other types of feedback other than simple preferences and dislikes are possible, for example, buttons that represent the type of emotional response to a track, such as happiness, sadness, or chilliness . The feedback mechanism need not be a button, but can be a knob, dial, thumbwheel, touch screen, voice recognition, biometric reader, or any other device that can receive user input. The shape of the button can be different from the shape shown in the figure.

  The shape, contour, position, and other aspects of the feedback device can be selected to give the user an intuitive tactile sensation that corresponds to the function of the button. For example, by placing the + button on the-button and making the plus button larger than the-button, it is possible to convey an intuitive sense of the meaning of each button. The surface contours of two different preference buttons can be arranged to show their meaning intuitively. For example, the + button can have a concave surface (usually felt as comfortable) and the-button can have a convex surface (usually felt as uncomfortable).

  Place preference buttons close to each other, near one or more primary transport buttons, for example, to allow the user to press any of these buttons with just a finger movement without movement of the entire hand It is also desirable to place them close enough. As can be seen from FIG. 24, one useful form of clustering buttons is to place + and-buttons on top of each other and the front and back transport buttons on the left and right of the vertical axis connecting the + and-buttons. Put it right next to it. For example, when attached to a steering wheel, the user can use the thumb to rate and transport without moving the entire hand. Besides being convenient, such an arrangement may be safer than other arrangements for use in a vehicle.

  The system can provide feedback to the user regarding the expressed preferences and other matters. In one approach, feedback can be provided on embedded screen 2301, as shown in FIG. In another approach, feedback can be provided on the LCD display 2501, as shown in FIG. Feedback may include identifying the played track, the user station playing, the length of the track, the elapsed time, and the system's interpretation of the user's likely preference level regarding the track being played. it can.

  In some embodiments, the buttons and displays described above can be augmented or replaced by a speech recognition system. The system provides verbal feedback to the user and listens to verbal commands. Instead, the system can provide verbal menu selection, which can be selected by a button press or voice response.

In an embodiment that provides feedback to the user using a display on-screen display, when the system is playing a stored track, the system will play the currently playing track, as shown in FIG. Information about can be displayed. The system uses current album name 2601, current track title 2602, artist name 2603, genre 2604, current track elapsed time 2605, currently playing user station 2606 (see "User Stations and Presets"). ), And the source 2607 of the track can be displayed.

  In the search mode, as shown in FIG. 9, the system may display a menu that allows the user to search by genre 904, artist 903, track 902, or album 901. The system can display a range 905 of current search results, which can initially be the entire set of stored tracks. In some embodiments, the initial scope of the search is based on the currently generated track. The user can select similar tracks relatively quickly, for example, by the same artist, within the same album, within the same genre.

  When the user desires to edit a preset or user station, the system can display the screen shown in FIG. If the preset is not a user station, the user can add or remove a given track from the preset. If the preset is a user station, the user can give explicit positive or negative feedback about the station.

  When a smaller display is required, such as an automobile or portable stereo system, the display can take the form of FIG. The system continuously displays the current mode 2701 and elapsed time 2702. Screen information 2703 for the current artist (eg, “The Beatles”) and current track title 2704 (eg, “Hard Day's Night”) may also be scrolled.

  When the user presses an explicit feedback button (eg, button 1814 or 1815 in FIG. 18), any embodiment can confirm feedback on the display (eg, “Feedback acknowledged”, etc.) By displaying automatically). In some embodiments, confirmation is displayed upon receipt of implicit feedback as well as explicit feedback.

  As the system is playing a track, the display can show an indicator 2608 (FIG. 26) that represents the system's understanding of the user's preference level for that track. This indicator can be, for example, one of the five icons shown in FIG. The five icons can include a solid plus 3101, a contour plus 3102, a circle 3103, a contour minus 3104, and a solid minus 3105. Other icons, other numbers of icons, and other meanings of icons can also be used.

  When the user enters the setting mode, a setting screen 2801 is displayed (see FIG. 28). Using this screen, the user can change the preset or station 2802 currently being played. Using this screen, the user can configure the system to automatically turn off 2803 after a time specified by the user or to turn on at a specified time. This screen allows the user to adjust standard audio settings 2804 such as treble, bass, and balance. Using this screen, the user can display and modify advanced settings 2805.

  When the user selects editing of advanced settings, a screen such as 2806 can be displayed. The system displays the currently playing station 2807, displays information about the number of tracks 2808 with feedback recorded and the content of the feedback 2809, and displays the artist 2810 with the highest feedback rating at the current station. It is also possible to allow the user to clear 2811 all information for the current station.

  Similar information and options can be presented verbally to the user when the system uses speech synthesis rather than an on-screen display.

In terms of user stations and preset concepts, a preset represents information about each stored track that affects the system selection of the track to be played. The mechanism by which the system uses this information depends on the current mode, and some preset information is applicable only to certain modes. For example, when in track mode, bin mode, or virtual CD changer mode, the preset contains simple yes / no data for each track or album that indicates whether the track or album is included in the current preset. included. In user station mode, the preset includes information used by the system to select a track in the automatic selection mode. For example, in FIG. 12, a feedback record 1213 can be included for each node in the stored track graph (see “Metadata”).

  The selection system can maintain any number of preset sets (see database schema in FIG. 15). For example, the system can maintain four presets, each of which is represented by a physical user interface button or other user interface item labeled 1 to 4 (FIG. 23). (See item 1805). While a track is playing in either mode, the user may specify that the track should be added to a given preset, for example by holding down the corresponding button. it can. When the system is in automatic selection mode, this has the same effect as giving a positive rating to the current track. The user activates the preset by pressing the corresponding button. Each preset can be assigned a number or name for display and selection.

  As can be seen from FIG. 15, the system can maintain at least three types of presets. One is a type that allows the user to create individual playlists by assigning individual tracks to presets. The second type allows a user to assign a CD or group of CDs to a preset and play tracks from the CD. A third type of preset is referred to as a user station. In addition, the user can set a CD preset to play a single CD.

  The user station is a preset used in the automatic selection mode. Each user station associates a probability with each track stored in the system. This probability is generated by the automatic selection algorithm described below and is dynamically updated based on user-indicated preferences, for example, implicit, explicit, or a combination of both. While a preset is in use, the selection of tracks to be played is based on a specific set of probabilities associated with the entire collection of tracks stored in the preset. Multiple presets can be defined as playing tracks that are selected based on different probabilities of items across the collection.

  In some embodiments, user stations can be associated with individual users. Thus, if there are three users of the system, user station 1 can reflect user 1's preferences, user station 2 can reflect user 2's preferences, and user station 3 can reflect user 3's preferences. Can be reflected. The user station 1 plays the selection based on the scoring of the track derived from the user preferences shown during the playback of the track by the user station 1, and so on.

  Some embodiments may allow automatic user identification and automatic use of user presets. This can be achieved by biometric identity verification. For example, the system can be equipped with a voice recognition function and the user can be identified by the sound of the user's voice. Once identified, the system can automatically switch to the user's station. Other systems can use fingerprint identification, retinal scanning, or other biometric measurement techniques. Still other systems can interface with external identification devices. For example, a system incorporated in an automobile can switch to the user 1 station when the user changes the driver's seat to the preferred location stored before the user 1.

  In some cases, rather than automatically updating the preset each time the system receives feedback from the user, the preset can be updated only upon explicit instruction from the user. In that way, the user can return to the preferred preset even after the session in which the preset was used to select the sequence of playback items.

The module diagram 21 represents the functional modules included in one possible embodiment and the nature of their interaction. In this embodiment, the MusicPlayer 2107 controls the playback of the track in response to the input, and supplies status information regarding the playback of the track to the PlayModes module 2103 in the form of an End Of Track Playback message and a TimeUpdate message 2122. The MusicPlayer 2107 sends a TimeUpdate message and a NewTrackPlaying message 2118 to the on-screen display (OSD) module 2101. The MusicPlayer 2107 sends an End Of Track Playback message 2115 to the PlayerStates module 2104. When the MusicPlayer 2107 sends a RipTimeUpdate message, a CDInserted message, a CDEjected message, and a CDReady message 2116 to the ripper module 2106 (see below), the CD is inserted into the ripper module for the time remaining in the ripping process. Informs when the CD is removed and when the CD is ready for storage.

  The ripper 2106 responds to the store command from the control user interface module 2102 and sends a message 2117 to the autoselect module to add the stored track to the stored track graph (see “Metadata”). The ripper 2106 asks the automatic selection module whether the given track is already stored in the graph (2117). The ripper module sends a ripTrack message 2123 to the MusicPlayer module to instruct the MusicPlayer module to actually store the track. The MusicPlayer sends a confirmation message when the storage is complete.

  The PlayModes module 2103 responds to a button press 2109 from the control user interface. The PlayModes module 2103 sends a message 2111 to the PlayerStates module 2104 to change the current state to “stopped” or “playing”. The PlayModes module 2103 queries (2124) the auto selection module to find out which track to play next or obtain the current playlist. The PlayModes module 2103 can also instruct (2124) the automatic selection module to process the feedback event. The PlayModes module 2103 also maintains track history, user station state (see “User Stations and Presets”), and virtual CD changer state (see “Modes of Operation”).

  The PlayerStates module 2104 responds to transport control events from user control and commands from the PlayModes module. The PlayerStates module 2104 also serves as an interface between the remote control and the MusicPlayer module. The PlayerStates module 2104 passes the transport control instruction and the track playback instruction 2112 received from the control user interface and the PlayModes module to the MusicPlayer module. The PlayerStates module 2104 can query (2113) the PlayModes module for the current track, previous track, and next track.

  An on-screen display (OSD) module 2101 represents information about what is currently being played by the user. The OSD module 2101 can handle user interaction to “reset” or otherwise edit the preset and send a reset message 2120 to the auto-selection module. The OSD module 2101 also displays the user interface of the search interface and sends the search result 2121 to the PlayModes module. The OSD module 2101 queries the automatic selection module 2119 to retrieve the search results (search parameters supplied by the user are provided to the OSD module). The OSD module 2101 also queries (2119) the automatic selection module for information related to the currently playing track and presents it to the user.

  The automatic selection module 2105 selects a track to play when the unit is in automatic selection mode (see “Modes of Operation”). The automatic selection module 2105 maintains all feedback information for all user stations (see “User Stations and Presets”). The automatic selection module 2105 performs a search when requested by the user via the OSD module, and converts the query formulated by the OSD module into a list of tracks to be played. The automatic selection module 2105 supplies information related to the currently playing track to the OSD module. The automatic selection module 2105 requests the user interface to edit the stored track graph data. This module maintains a stored track graph, so when the system is in virtual CD changer mode, track mode, or bin mode, the PlayModes module automatically checks to see the next played track from this data structure. The selection module must be queried 2124.

  The Control User Interface (ControlUI) 2102 module dispatches user events. The Control UI 2102 sends a button press 2108 to the OSD module for user interaction, eg, in search mode. The Control UI 2102 sends a source change message, a mode change message, a rating message, and a next track message 2109 to the PlayModes module. The Control UI 2102 sends a transport control message 2110 excluding the next track message to the PlayerStates module. Control UI 2102 sends a memory button press 2114 to the Ripper module.

The track storage auto-select mode, track mode, bin mode, virtual CD changer mode, and search mode all require that track data be stored on the system hard disk. This is done by the ripper module, which operates as shown in FIG.

  The module is in a “no disk” state 501 when no disk is inserted (and no other source is operating). When a disc is inserted (or when a user begins playback from another source, such as a radio or cassette tape), the system reads Table of Contents (TOC) data when available ( (See “Data Storage” below), and enters the “Disk Present / Idle” state 502. If there are more tracks to store when the user presses the store button, the system saves a list of unstored tracks and enters a “stored” state 509. If there are no more tracks to store 508, the system displays an error message.

  While in storage, tracks are simultaneously recorded on the hard drive while other modules are processing playback of the source. When all tracks have been stored, the module can display 506 a confirmation message to the user and return to the idle state 502, but instead return to another source mode and play the source from the beginning. it can. If the source is removed before storing all tracks, the system returns to the no disk state 501. If an event occurs when the source is removed, the user presses stop, the user changes the source, the hard disk runs out of space, or the system is unable to continue storing the current track The stored data of the current track is deleted.

  In other embodiments, the track need not be played while being stored on the unit's hard drive. In that case, track storage can be completed in a shorter time. The storage module operates independently in these embodiments, and the user can continue to listen to tracks from any source, including the system hard disk, while the storage module is storing tracks.

  In other embodiments, “background storage” is allowed. This feature allows the user to indicate that a given track must be stored. The system later defers actual storage until the system is no longer in use. If the user uses the system while the system is storing the track in the background, the system aborts the storage function. If a user plays a track from a CD that is queued for background storage, that track can be stored as the user is playing.

  Background storage can be combined with standard storage functions as described below. Standard storage can be used normally, but when the hard disk runs out of space, the user removes the CD, or other events occur that make it impossible to store tracks, the system The track can be queued in the background store and stored later when it becomes possible.

  Some embodiments perform minimal processing during storage and postpone advanced functionality until the system is idle. In these embodiments, the length of time that the original medium is required is reduced, and most of the signal processing occurs during idle time.

Metadata metadata means information about a track other than the underlying track data itself. There are numerous examples of potential metadata sources and hierarchy types. For example, if the stored track is music in nature, the system can use a digital signal processing algorithm to analyze the amount of bass on a particular track, for example. The track can then be assigned a category such as “heavy bass”, “light bass” or between. Instead, if the track represents non-music data, the appropriate metadata for that type of data is used. For example, if a track represents a command in a computer program, the appropriate metadata category can be set to “scope” (eg, document level command, global command, selection level command), “type” (eg, font, style, Reference, layout, etc.), menu (eg, file, edit, display, etc.), and command name (eg, insert page, etc.). Like the music data, this metadata can be pre-packaged with the software, or it can be made available separately, for example via the Internet. The following section uses an example of stored music. The following description exemplifies the creation of a stored track graph suitable for representing characteristics specific to music data, such as artist, genre, and style. In other embodiments, different structures with different relationships between nodes are used depending on the type of data stored.

  Returning to the music example, the metadata has a structure such as that shown in FIG. The individual tracks 3001 together constitute an album 3002, a plurality of albums constitute a work of an artist 3003, and a plurality of artists constitute a genre 3004. In many embodiments, additional layers are added as described below.

The automatic selection mode can be utilized, for example, at any of the following three levels of metadata support:
(1) No metadata.
(2) includes a list of “track titles” indexed by a unique album identifier and potentially ordered by the columns “genre”, “artist name”, “album title”, and “track number” A database of metadata supplied externally. One source of such data is the Gracenote CDDB database, and all data having these fields are generically referred to herein as “CDDB data”. Please refer to FIG.
(3) includes an album table and is indexed by a unique album identifier and potentially a list of columns “genre”, “style”, unique artist identifier, “album title”, “sortable album title”, and “ An externally supplied database containing a list of “similar” albums (referenced by a unique ID). This database is indexed by a unique artist identifier and potentially lists columns “Artist Name”, “Sortable Artist Name”, “Style”, and “Similar” Artists (referenced by unique ID) The artist table is also included. The database also includes a static “style” table, where each style is linked to a “genre”. This database is indexed by a unique album identifier, the columns “Media Index” (disc serial number of the multi-disc set), “Track Number”, “Track Title”, “Sortable Track Title”, and “Artist Unique” A “track” table with “ID” is also included. One source of such data is the All-Music Guide. In this specification, data having these fields is generically referred to as “AMG data”. Please refer to FIG.

  Other embodiments of the selection system can handle all metadata formats that can be converted to the internal graph format of the selection system. The “stored track graph” of the system is a graph with nodes of different levels of generality. The nodes can be connected by an ancestor-descendant relationship or by a cousin-cousion relationship, as shown in FIG. The level of generality, and hence the ancestor-descendant relationship, is transitive.

  The data structure of the system's stored track graph is summarized in FIG. Each item of metadata forms a node in the half-layer network. The figure shows a number of different types of nodes, each derived from a generic node object. The node object 1209 (and thus all nodes) has a “name” (used when something must be displayed to the user), “id” (unique identifier), “sortname” (name sortable) Version), and a set of feedback records 1214. The feedback record consists of one record maintained globally and one record 1213 per user station, in which all feedback, score, and playback history information is stored.

  Hierarchies are represented in this figure by vertical dimensions and internode connections. A connection (ie, an ancestor-descendant relationship) from one level of node A to a higher level of node B is based on whether the entity represented by node A is part of the entity represented by node B, or Implies that it has been executed.

  Reading from the bottom of the figure, the track node 1201 includes a standard node field, trackNumber (position on the CD), artist and genre (when metadata is available), A field for its duration is included. Each track node is connected 1222 to a single CD node 1202 and tracks from the multi-artist CD are connected 1225 to one or more artist nodes 1204 (when this metadata is available).

  The CD node 1202 stores a table of contents (TOC) string of the disc and an ordinal “media index” (when available) representing the disc number of the multi-disc set. Each CD node is connected to a single album node 1203.

  Album node 1203 represents either a single CD or a multiple disc set. The album node 1203 is connected 1224 to a single artist node 1204 when the album includes a single artist track, and is not connected to any artist node when the album includes multiple artist tracks. Instead, the track node 1201 is directly connected 1225 to the artist node 1204. An album node can be connected 1227 to zero or more style nodes 1206. An album node can also be connected 1221 to zero or more “similar” albums by cousin-cousin links.

  Artist node 1204 represents a single performer, which is defined as one or more individuals who have recorded one or more tracks. Artist node 1204 can be connected 1226 to zero or more style nodes 1206. An artist node can also be connected 1229 to zero or more “related” artists via a cousin-cousin link.

  A style node 1206 represents a music sub-genre. The style node 1206 is connected 1205 to at least one genre node 1207. When the metadata does not provide both style information and genre information, the genre can be determined from the style by a static lookup table (see 1101 in FIG. 11). For example, the style “Classic Rock” is mapped to the genre “Rock”.

  The genre node 1207 corresponds to a wide class of music (eg, “rock”, “classic”, “jazz”). Genre node 1207 is connected 1228 to a root node (“Everything” node) 1208.

The root node 1208 is an ancestor of all other nodes. A root node 1208 represents the entire music collection.
Connections between nodes are represented by data structure 1229. Connections are bi-directional (and thus include parent or left cousin 1230 and child or right cousin 1231), each connection has a numeric weight ("link weight") 1232 that represents the strength of the relationship, and the nature of the relationship Has a type 1233 for designating. The link weights can be derived from the external metadata information, can be static, or can all be equal to one.

  If explicit metadata is not available, the system does not have a direct way of classifying tracks. For example, when a new music CD is purchased and loaded into the system, the system knows the relationship between the tracks on the CD, but does not know the genre or style of each track. Therefore, a node cannot be created at an appropriate position in the stored track graph. In some embodiments, this problem is addressed by prompting the user for the required information (eg, using speech synthesis) and using the results to build metadata in the system. . For example, the system can ask "What genre is this?" The spoken result “jazz” can be recognized and stored.

There are five events that the user feedback type selection system learns from the user and uses to determine user preferences (the five events are described, but other types of events are possible).
(1) The user presses a button or otherwise gives an explicit positive response at some level (eg, “Rate +”);
(2) The user presses a button or otherwise exhibits an explicit negative response at some level (eg, “Rate-”);
(3) For example, if the user presses the button on the assumption that the button is pressed within a predetermined time period (for example, more than 2 seconds and less than 2 minutes) after the track starts to play, In other ways, indicates that the system should skip to the next track, or otherwise aborts the current track ("next track");
(4) that there was some type of user interaction within a set time period, eg one hour centered at the beginning of the track, indicating that the user is likely to listen to the track and complete it deliberately Assuming the item is played from beginning to end ("track complete successfully");
(5) The user raises the volume (“Volume +”).

  The first two items are explicit user actions that tell the system something about preferences. The remaining items represent only implicit information about the user's preferences. Other types of feedback about preferences, both explicit and implicit, can be used.

In the node scoring item selection system, a score based on feedback from a user is assigned to a node of a stored track graph. Because of the connection between the nodes of the graph, feedback applied to one node (eg, user response to a track) potentially affects all nodes of the graph. Thus, the system can make deductions about the user's response to each stored track based on the user's response to only one track. In addition, the system can store short term preferences and long term preferences. This feature allows the system to quickly adjust to changes in the user's mood without losing information about the user's general long-term track preferences.

  FIG. 31 shows an initial distribution of scores between tracks prior to receiving feedback and two track score distributions representing an example of the desired final distribution. Initially (3106), the system has not received user feedback, so the user is assumed to be neutral 3103 for all tracks. When explicit and implicit feedback is received, the system deducts a track that is similar to the track for which explicit feedback was received. Ultimately, one goal of the system can be to achieve a double bell curve distribution 3107. The user provides either explicit positive 3101, explicit negative 3105, implicit positive 3102, or implicit negative 3104 feedback for a small number of tracks. Using that information, the system deducts the user's preferences and assigns positive or negative scores to the remaining tracks based on the relationship of the unrated tracks with respect to the rated tracks.

  As described above, the selection system maintains a graph representing all tracks, metadata about that track, and relationships between that track and other tracks. As described in detail below, the system uses this graph structure to propagate feedback to nodes other than the currently playing track. In addition, as fully described below, a node can optionally track long-term preferences separately from short-term preferences.

Long-term preferences can be tracked as follows, but the numerical weights are only provided as one possible example (see FIG. 13).
(1) When the user presses the plus button 1306, the system sends a +1 feedback event to the track node 1311 (1301).
(2) When the user presses the minus button 1307, the system sends a -1 feedback event to the track node (1302).
(3) When the user presses the next track button 1308, the system sends a -1/3 feedback event to the track node (1303).
(4) At 1310 when the track is successfully completed, the system sends a 1/3 feedback event to the track node (1305).
(5) When the user presses the increase volume button 1309, the system optionally sends a +1/6 feedback event to the track node (1304).

  These feedback events modify the accumulated Score variable 1210 and the accumulatedWeight variable 1211 of the track node 1201 and all its ancestors in the metadata network. accumulatedScore is equal to the sum (positive and negative) of all feedback for that node, and accumulatedWeight is equal to the sum of the absolute values of each feedback event for that node. Therefore, the ratio accumulatedScore / accumulatedWeight is always in the range [-1, +1]. This ratio is referred to as the local score. If the local score is +1, this means that all feedback received has been positive. If the local score is -1, this means that all feedback received was negative. If accumulatedWeight is 0, the local score is defined as 0.

  In some embodiments, accumulatedWeight and accumulatedScore are bounded. If accumulatedWeight exceeds a fixed boundary value, both variables are multiplied by the ratio fixedBoundaryValue / accumulatedWeight, thereby keeping them within the specified boundary while maintaining the ratio of these variables.

  Short-term mood-based preferences can be handled as well. For example, if a user normally enjoys rock music but now wants to listen to classical music, the system receives the user's short-term preferences and adjusts them without substantially modifying the long-term preferences. . Thus, in this example, the system quickly interprets the user's feedback as meaning that the user only wants to listen to classical music at this time, but when the system is later turned on, the system Remember that rock enjoys the most.

  For this reason, instead of affecting accumulatedScore and accumulatedWeight, the affected variable will be modBonusScore1212. In this configuration, when the feedback event is about to be processed, all the modBonusScore variables across all nodes are reduced toward 0 (eg, by multiplying each value by a fraction such as 0.9). Next, the system adjusts moodBonusScore for each track that receives feedback and its ancestors, for example, by adding +0.1 for positive feedback and -0.1 for negative feedback. The modBonusScore can be bounded and in the above example it is guaranteed that the modBonusScore is in the range [-1, +1].

A system that maintains both long-term preference data and short-term preference data can function as follows.
(1) The “Rate +” event results in a large positive change to the moodBonusScore and a small positive change to the accumulated variable.
(2) A “Rate-” event results in a large negative change to the modBonusScore and a small negative change to the accumulated variable.
(3) The “Next Track” event results in a small negative change to the moodBonusScore and a small negative change to the accumulated variable.
(4) The “track completed successfully” event results in a small positive change to the moodBonusScore and a small positive change to the accumulated variable.
(5) The “Volume +” event results in a small positive change to the moodBonusScore and a small positive change to the accumulated variable.

Thus, the accumulated variable has infinite memory, and the modBonusScore variable has limited memory (since it decreases towards 0).
Further, in some embodiments, the system allows seeding of scores from the search screen (see “Modes of Operation”). That is, the user can locate the track on the search screen and press one of the explicit rating buttons to apply feedback to the node and its parent.

  In some embodiments, pressing the “Rate-” button stops playback of the current track and starts the next track. For the user, this is the same effect as pressing the “Next Track” button, but this has a stronger impact on the immediate behavior of the system.

  In other embodiments, in addition to skipping the current track as described above, there is a short pause before the playback of each new track. The name of the track being played is displayed on the on-screen display. The user can immediately skip to the next track by pressing the “Rate-” button before playing the track. In this case, the user provides explicit feedback without playing the track.

  Each node in the graph maintains its own score related data. Each node includes a set of feedback records 1214 (see FIG. 12). This set includes one group of variables and an additional global group for each preset. Each group includes three numeric variables: accumulatedScore, accumulatedWeight, and moodBonusScore.

  Each track node, CD node, and artist node also has a single integer value variable lastPlayedAtIndex. If lastPlayedAtIndex holds the value -1, the track represented by that node has never been played by the system. The system maintains a global circular playback counter that starts at 0 and is incremented as each track is played and returns to 0 when a specified value, eg, 1024, is reached. When a track is played, all nodes whose lastPlayedAtIndex variable is equal to this global counter value are reset to “−1”. For example, if the 5th track played by the system is X and the system is currently playing the 1029th track, the global counter value is 5 (1029 mod 1024 = 5), so track X LastPlayedAtIndex of 5 is reset to -1. Thus, the system “forgets” that this track has been played. Next, the current value of the global counter is set to the value of lastPlayedAtIndex of the current track node, CD node, and artist node. At this point, the expression (globalCounter-lastPlayedAtIndex) mod 1024 represents how recently the track node, CD node, or artist node was played, assuming that lastPlayedAtIndex is greater than -1.

  The system also maintains a first-in first-out list (see FIG. 14) that represents the most recently played track for each preset. This playback history is used only when the user moves back and forth in the playback history using transport control. Thus, if the system is currently playing track 1402 and the user presses the previous track button, the system will play track 1401. The next track that the system plays is track 1402. When the system completes playback of track 1402, a new track is selected and added to the end 1404 of the list, and the oldest track 1403 is removed from the list.

  If the user presses the previous or next track and a track that would otherwise have been played is deleted from the hard disk, the system will skip that track and immediately before or immediately after it in the list Play tracks.

Data storage format tracks are recorded from a CD to a hard disk or other random access storage medium. If a track represents music data, each track is potentially a separate file, potentially in the MPEG-1 Layer 3 format (“MP3”; a standard defined by the Moving Picture Expert Group), AAC format (Advanced Audio). Coding, also a standard defined by the Moving Picture Expert Group), or other suitable format and stored on the medium. Along with each track, table of contents information is also stored if available on the original medium. The track index on the CD is also stored. The particular format of the new media is not important as long as track numbers and table of contents are available. In some embodiments, a separate directory is created for each album (see FIG. 22). These directories are numbered sequentially, such as CD0001, CD0002, for example. Within the directory, track files are numbered sequentially (TR001, TR002, etc.). The table of contents can be stored, for example, as a character string in the ID3 header of each MP3 file.

  As tracks are stored on the hard drive, nodes are added to the stored track graph. As noted earlier, the selection system can obtain metadata from multiple sources to build the graph. In some cases where the track is music, the selection system retrieves data provided by All Media Guide (AMG) and data provided by Gracenote's CDDB service. First, the system retrieves the stored track graph of the CD node whose TOC matches the current CD's table of contents. If the CD does not exist, the system queries the AMG database using the table of contents of the CD as an index. If the AMG database returns a match, the system adds the CD node, and if it does not already exist, adds the album node using the data from the AMG database. Note that all of its variables are set to 0 whenever a new node is created. A track node is created and added to the network under the newly added CD node. The system uses the list of similar albums in the AMG database to connect the album node to other albums in the network. A track node is connected to the album node. The album node is connected to the appropriate style node, and the style node is created if it does not already exist.

  If this is a single artist CD and the album node did not previously exist, or if it is a multiple artist CD, the system creates an artist node if it does not already exist. An artist node is connected to an album node if it is not a multiple artist CD. If the artist did not previously exist, the artist node is connected to the style node specified in the artist entry in the artist table of the AMG database. In the case of multiple artist CDs, the system connects the artist node directly to the track node, otherwise it connects to the album node.

Style nodes that did not exist before are connected to the appropriate genre node (created as needed), and the new genre node is connected to the root node.
If no metadata is returned, the system can query the CDDB database. If there is a match, the node is created as described for the AMG database. If there is still no match, a track node and a CD node are created, along with dummy “unknown” nodes for album, artist, and style levels.

  Alternatively, the CDDB database can be queried first. If a match is found, the AMG database is queried second, and if available, more detailed information from the AMG database is used to augment the CDDB data.

  As noted earlier, CDDB data and AMG data are only two possible types of metadata and are particularly suitable for music tracks. When other types of data are used, other hierarchies with different levels are created to map the information made available by the appropriate metadata to the data. For example, if the database consists of issued patents, these metadata fields have been described above if the metadata provided is the field, the inventor, the prior art described, and the name of the invention A stored track graph is created that maps to the structure.

Propagation of feedback User feedback can be applied to a given node as follows. The system calls the function insertFeedbackAtNode to apply the feedback (ie, the track node that is being played), an integer preset number, a modBonus (representing a short-term score change), and floating-point rawScore delta and weight delta (long-term (Represents the change in feedback). First, the selected node is initialized.

node. currentFeedbackFlag 1218 = 1
node. currentFeedbackRawScore 1216 = rawScore
node. currentFeedbackWeight 1217 = weight
node. currentMoodBonus 1219 = modBonus
The system then traverses the graph to go all the nodes at the lower level before the upper node. When going to each node, if the currentFeedbackFlag of that node is set, consider the cousin link of each node. Initialize the temporary variables tbonus, tweight and trawScore.

tweight = node. currentFeedbackWeight * link. weight * COUSIN_PENALTY
trawScore = node. currentFeedbackRawScore * link. weight * COUSIN_PENALTY * MAX_IMPLICIT_CONTRIBUTION
tbonus = node. currentMoodBonus * link. weight * COUSIN_PENALTY
If the cousin node's currentFeedbackFlag is not set and tweight is greater than the value of the cousin node's currentFeedbackWeight, set the values of the currentFeedbackRawScore of the cousin node to the values of the trascore and currentFeedbackHeight. In one embodiment, COUSIN_PENALTY is set to 0.4, and MAX_IMPLICIT_CONTRIBUTION is set to 0.8. If the absolute value of the coordinator node's currentMoodBonus is smaller than the absolute value of tbonus, tbonus is set in the cousin node's currentMoodBonus.

  Next, each node's parent link is considered in turn, continuing the assumption that the current node's currentFeedbackFlag is set. Initialize the temporary variables tbonus, tweight and trawScore.

tbonus = node. currentMoodBonus * link. weight
tweight = node. currentFeedbackWeight * link. weight
trawScore = node. currentFeedbackRawScore * link. weight
When the parent node currentFeedweightWeight value is smaller than tweight, the parent currentFeedbackFlag is set to 1, and the parent currentFeedBackRawScore value is set to the trascore and currentFeedweightWeight values. If the absolute value of the parent node's currentMoodBonus is less than the absolute value of tbonus, set it equal to tbonus.

The system repeats this process for the parent of the current node.
After visiting all nodes, visit all nodes in reverse order (starting from Everything node and going down to the track node). When visiting each node a second time, if currentFeedbackWeight is non-zero, adjust the feedback variable (of the specified preset) for that node as follows:

accumulatedRawScore + = currentFeedbackRawScore
accumulatedWeight + = currentFeedbackWeight
Each node type can have a weight limit. One example is shown, although other weights are possible.

FeedbackLimit = [4.0, # track
10.0, # CD
10.0, # Album
20.0, # Artist
40.0, # style
60.0, # Genre
100.0] # If accumulatedWeight is greater than the limit for that type of node, set AccumulatedWeight to be equal to the limit. If the absolute value of accumulatedRawScore is greater than the limit of that type of node defined above, normalize to equal the limit.

  Also, (1-modbon) is applied to the node's moodBonusScore (of the specified preset), and the value of currentMoodBonus is added. Finally, the values of currentFeedbackFlag, currentFeedbackRawScore, and currentFeedbackWeight are all set to zero. At this point, the system is propagating feedback to all ancestors and the first cousin of the ancestor of the selected node.

  In one embodiment, the insertFeedbackAtNode function is called for only one of the four possible combinations of rawScore, weight, and moodBonus.

1. When the “Rate +” button is pressed, rawScore = 1.0, weight = 1.0, and modBonus = 0.1
2. When the “Rate-” button is pressed, rawScore = −1.0, weight = 1.0, and modBonus = 0.1
3. RawScore = -MAX_IMPLICIT_CONTRIBUTION * IMPLICIT_PENALTY, weight = IMPLICIT_PENALTY, and moodBonus = 0.05 when the “Next Track” button is pressed within the appropriate time window
4). When the track playback is completed, rawScore = MAX_IMPLICIT_CONTRIBUTION * IMPLICIT_PENALTY, weight = IMPLICIT_PENALTY, and moodBonus = 0.05
In one embodiment, the constants are set as follows: MAX_IMPLICIT_CONTRIBUTION = 0.8 and IMPLICIT_PENALTY = 0.33.

  Once the feedback has been propagated to the graph, the node scores need to be calculated taking into account both short-term and long-term preferences. This includes top-down traversal of another graph, but this step can be combined with the previous traverse. First, starting from the Everything node and going down, the system zeros the rawScore 1216 and weight 1217 of the node. For each current node parent (if any), add the preset parent node's rawScore (value multiplied by the link weight) to rawScore, and add the link's parent node weight weighted to the weight to the weight To do. After all parent feedback is added, both rawScore and weight are multiplied by PARENT_PENALTY. In an example embodiment, PARENT_PENALTY may have a value of 0.2. Next, the system adds the current preset value of the accumulatedRawScore node to rawScore, and adds the accumulatedWeight value to weight.

  The system maintains a list of constants, MOOD_WEIGHTS [GRAPH_LEVELS], and there is one mood weight constant (MOOD_WEIGHTS) for each level of hierarchy in the graph. In one embodiment, each MOOD_WEIGHT is 0, except that the CD level MOOD_WEIGHT is 1, the artist level MOOD_WEIGHT is 1, and the style level MOOD_WEIGHT is 2.

  The selection system multiplies the moodBonusScore of the current node by that node level mood weight and adds this value to rawScore. The system adds the current level of mood weight to weight. The system repeats this process for each child of the current node.

  The ratio rawScore / weight is defined as the score of the specified preset node, unless weight is 0, and when weight is 0, the score is defined as 0. Note that a node's score depends on the ratio of all scores of its ancestors in the network. Since all nodes come from Everything nodes, this implies that all feedback events affect the scores of all nodes.

When the next track selection system is in "auto-select" mode, the system selects the next track to play based on the current risk tolerance and each node's score combined with several other variables To do. Thus, in some embodiments, the track selected for playback may not be the highest rated track. Every track has a certain probability of being played in each selection cycle, even the lowest score track, and the exact probability depends on the current risk tolerance of the system. The risk tolerance variable keeps the system from making two mistakes at the same time, while at the same time trying to achieve a degree of variation that is not possible when the system is playing the highest rated track. When receiving negative feedback, the system becomes more conservative (conservative) to avoid the possibility of repeated mistakes. When receiving positive feedback, it gradually becomes less conservative.

  In some embodiments, the selection process proceeds as follows. First, the system counts artist nodes, CD nodes, and track nodes that have scores that exceed a constant threshold. In one embodiment, the threshold is 0.2. As a result, the system knows the number of artists, CDs, and tracks that the user “likes”. These counts are assigned to the variables artistCount, CDCount, and trackCount, respectively.

  Each preset has a conservative throttle variable 1215 that determines the amount of risk that the system takes when selecting a track with a relatively low score. The conservatismThrottle is bounded, for example in the range [0.5, 7.5] and is adjusted whenever a feedback event occurs. If any of these adjustments causes the throttle to exceed a threshold, the throttle variable is appropriately set to a maximum or minimum value.

(1) When the “Rate-” key is pressed, the throttle is multiplied by 10.0, which has the effect of reducing the risk taken in selecting the next track, so that the user is already at least conservative. It reflects the notion that it is desirable not to take too high a risk in a dissatisfied context.
(2) When the “Next Track” button is pressed within the appropriate time window, the throttle is multiplied by 1.5, which is not as great as when the rate-key is pressed, but there is a risk of being taken decrease.
(3) When the track is successfully completed, the throttle is multiplied by 0.9, thus increasing the risk taken in view of the success of the previous track selection.
(4) Finally, when the system selects a track with a score less than a constant threshold (eg 0.2), multiply the throttle by 1.51 to increase the risk of being taken by a larger amount.

  Note that “Rate +” and “Volume +” do not affect the throttle.

In order to select a track for playback, the temporary variable totalLikelihood is set to zero. Next, visit each track in any order. As described above, the raw score / weight is set in the variable score. Next, when the lastPlayedAtIndex of the track is not equal to −1, this is converted into a distance as follows (the% symbol represents a remainder operation)
distance = (globalPlayIndex-lastPlayedAtIndex)% 1024

A penalty is applied to score if the distance is less than the track count (calculated above).
score- = 2.0 * (1.0-trackDistance / trackCount)

Similarly, the CD node and artist node of a track are examined and a further penalty is applied if the distance is less than the CD count or artist count.
score- = 0.6 * (1.0-CDDistance / CDCount)
score- = 0.4 * (1.0-artist distance / artist count)

  At this point, the score for a given node represents the average feedback received by that node that is penalized if a track, CD, or artist was recently played. Here, score is converted into a likelihood value (likelihood) by the following equation, where pow (x, y) represents x to the power of y.

likelihood = pow (10.0, score * conservatismThrottle)
The exact form of this function is not a problem, but when Score1> Score2, the ratio of two likelihoods of two different scores, L1 / L2, must increase for large values of conservatismThrottle and decrease for small values. That is, this function must exhibit an inflatable nonlinearity. Next, likelihood is added to the local variable totalLikelihood to generate a random number in the range [0, 1]. If the product of totalLikelihood and random number is less than partialLikelihood, the track under consideration is tracked with the local variable currentBestChoice.

  This process is repeated for each track. At the end of this process, the system selects the track represented by currentBestChoice for playback. One track is selected, and the probability that each track is selected is guaranteed to be equal to the track's likelihood value divided by the sum of all track nodes' likelihood values.

After the proposed user for future purchases has used the selection system for a sufficient period of time, the scoring values of the nodes in the graph correlate well with how much the user likes different groups of tracks. A score value close to +1 indicates a strong preference, and a score value close to -1 indicates a strong disgust.

  The system can be extended to temporarily insert data about CDs that the user does not actually own into the graph. That information can be derived from one of the previously described external metadata sources. Once inserted into the graph, the CD score can be calculated using the values inherited from the parent node of the CD. If the CD gets a high score, the system can recommend the user to buy the CD. After recommendation, the CD data is removed from the system graph.

The item suggestion selection system to be deleted uses a storage mechanism with a finite capacity. At some point, you are likely to run out of system storage capacity. When this happens, the system can list the tracks or CDs with the lowest score / weight values and suggest deleting them.

The pre-seeding system may allow stored track graph feedback to be “pre-seeded” by the user using preference data. For example, during the search mode, the user can indicate to the system that he likes or dislikes the entire artist, the entire album, or the entire genre. In that case, this feedback can be applied directly to the appropriate nodes of the stored track graph. In this example, preseeding simply sets the initial value of the node, and this value continues to be updated as described above in the normal course of track playback.

The smart state recovery user can switch to a mode other than the “automatic selection” mode for a short period of time. For example, the user can switch to the radio to look up traffic conditions or baseball scores. If the user returns to the “automatic selection” mode within a short period of time (eg, 1 hour), the system displays a state including the short-term preference variable, current track information, and playback history information in the “automatic selection” mode. Can be restored exactly the same as when it was interrupted. After this time period, the system behaves as soon as it starts by resetting the playlist history and short-term preference information.

When the on-screen display selection system is playing content from the hard drive (ie, modes other than CD mode and AM / FM mode), the on-screen display is defined by the state diagram of FIG. The OSD module is activated (1604) when playing a track from the system hard disk. In the “Now Playing” state 1601, minimal information can be displayed. When the user presses (1605) the “settings” button 1801 (see FIG. 18), more detailed information about the currently playing track is displayed or cleared (see FIG. 26). . When the user presses the enter button (1606), if the detailed information regarding the currently played item is shown, the preset edit screen of the preset edit state 1602 is displayed (see FIG. 19). ). On this screen, the user adds or deletes items from the preset (if the user station, gives the item positive or negative feedback), or clears the preset completely (including playlists and user stations) )be able to. When the user presses Enter (1607), the system asks the user for confirmation, and if the user presses yes, the OSD sends a reset message and returns to the playing state. If the user selects no or presses Exit or Settings instead of Enter, the system returns to the playing state without doing anything (1607).

  From the preset editing state or playing state, when the user presses “Search” 1812 (1609, 1610), the system enters search mode 1603 (1609) (see “Mode of Operation” above), Allow users to search for tracks to play. From search mode, when the user presses Exit or Search, the module returns to playing mode without doing anything (1608). If the user presses Play or Enter, the module resets the playlist and sets “PlayList” to the current playback state. The OSD module returns to the playing mode (1608). If the user presses the Settings button, the system displays detailed information about the item currently being played and returns to the playing state (1608).

Playback Status Module Some embodiments of the playback status module can operate as shown in FIG. This module enters the playing state 407 from the non-hard disk source 401 when the user presses Change or Smart Play or when a track is stored (403). If the system was last in play mode for less than k minutes (k is a constant, eg 5 minutes), this module plays the previous continuation, otherwise it starts from the beginning. When the end of the track is reached, or when the user presses track skip, the system proceeds to the next track if there are remaining tracks and returns to playing mode (406). If the user presses Previous track and the time exceeds the pre-specified time (eg 3 seconds) from the beginning of the track, or if the system is already playing the first track, Plays the current track from the beginning and returns to the playing mode (404). If the length of time that the system was playing is less than the threshold and there is a previous track, the system skips to the previous track and returns to playback (404). If the user presses Pause / Play during playback mode, the module switches to pause mode 416 (411) and pauses playback. If the end of the track has been reached and there are no more tracks to play, or if the user presses Stop, the module switches to stop mode (413).

  While playback is paused (416), if the user presses “Previous track” and playback proceeds beyond a pre-specified time (eg 3 seconds), or if the module is already the first track The module starts the current track from the beginning, but remains in pause mode (405). If the playback time is less than the threshold and there is a previous track, the module switches to the previous track and remains in pause mode (405). If the user presses “Skip Track” and there is a next track, the system skips to the next track and remains in pause mode (417). If the user presses “Pause / Play” while the module is paused, the module switches to the playing mode (411). If the user presses and keeps Previous Track, Next Track, Fast Forward, or Rewind while playback is paused, the module switches to silent FF / RW mode 429 (419). After a pre-specified timeout (eg, 5 minutes) or when the user presses stop, the module switches to stop mode (414). If the user presses “Play / Pause”, the system resumes playback (410).

  In the silent FF / RW mode 429, the module performs fast forward or rewind without sound. If the end of a track has been reached and there are more tracks to play (before the current track in case of rewind, after the current track in case of fast forward), the end of the next track or the previous track Skip to the beginning, and rewind or fast forward again with silence (426). This continues as long as the user holds the button pressed (427). If the end of the track has been reached and there are no more tracks to play, the module switches to “stop” mode (428). When the user releases the button, the module returns to the paused state (418). If the end of the track has been reached and there are no more tracks to play, the module enters stop mode 425 (428).

  In stop mode 425, if the user presses Next Track or Previous track and there is a next or previous track, the system skips to that track and remains in stop mode (424). If the user presses Play, the system starts playing the current track from the beginning (412). If the user presses and holds fast forward, rewind, next track, or previous track, the system enters FF / rewind mode 430 (423). If the user presses pause, the system sets the elapsed time to 0 and enters the pause mode for the current track (415).

  If the user presses and holds fast forward, rewind, previous track, or next track while in playback mode, the system switches to FF / rewind mode 430 (408). The module fast forwards or rewinds the track so that it can be heard. The module remains in this mode as long as the button is held (421). If the end of the track is reached and there are more tracks to play, the module switches to the end of the previous track for rewinding, to the beginning of the next track for fast forwarding, and continues the current state (420). When the user releases the button, the system switches to playback mode (409).

  When the source button is pressed (tape, AM / FM radio, or CD / DVD), the system switches to that mode after saving the current elapsed time of the track (402).

Variants Various embodiments have been described, but other embodiments are within the scope of the claims.

  For example, metadata that links the body of an item to be searched or used and the item hierarchically or otherwise can be targeted for similar selection based on the user's explicit or implicit preferences. For example, in the case of a database of items, the user can represent a normal search. The system can create the first “hit” from the search and display it to the user. The user can indicate satisfaction or dissatisfaction with the first hit. The system rescores items resulting from the initial search based on the relationships indicated by the metadata. The system can then deliver another hit to the user, similar to the next track selected in the system described for music. In the same way that the system described above adapts itself to deliver tracks of interest to the user, the database system adapts itself to deliver hits of interest to the database searcher. be able to. Using a specific example of such a database, patents can be searched in a patent database. Metadata that captures the name, assignee, patent number, classification, whether it is a prior art, design patent or utility model, inventor, and other information can be used. FIG. 34 shows the structure of an example of such a stored track tree.

  Another example is world wide web surfing. Metadata associated with the web page can include IP address, title, length, language, domain name, and other information. When the search makes a hit, the user can indicate whether it meets his needs. The system can quickly adapt to deliver more relevant hits to the user's browser.

  While several examples have been described above, other embodiments are within the scope of the claims.

It is a block diagram which shows a platform. It is a block diagram which shows a platform. It is a block diagram which shows a platform. It is a state diagram. It is a state diagram. It is a state diagram. It is a state diagram. It is a block diagram. It is a figure which shows a node. It is a figure which shows a display. It is a figure which shows a database schema. It is a figure which shows a database schema. It is a figure which shows a database schema. It is a figure which shows a database schema. It is a figure which shows a user interface. It is a figure which shows the history of a track | truck. It is a figure which shows a database structure. It is a state diagram. It is a flowchart. It is a figure which shows a remote controller. It is a figure which shows a display. It is a state diagram. It is a state diagram. It is a state diagram. It is a figure which shows a hierarchy. It is a figure which shows a user interface. It is a figure which shows a user interface. It is a figure which shows a user interface. It is a figure which shows a display. It is a figure which shows a display. It is a figure which shows a display. It is a figure which shows a graph. It is a figure which shows a hierarchy. It is a figure which shows a graph. It is a figure which shows a user interface. It is a block diagram. It is a figure which shows a hierarchy. It is a block diagram which shows a system. It is a block diagram which shows a system.

Explanation of symbols

10 System 12 Supply of Available Tracks 14 Track Source 16 Track Player 18 Next Track Identifier 20 User 22 Track Selector 24 Logic 26 Storage 28 Track Selection Module 30 User Interface 32 Input Side 34 Output Side 36 Track Scoring Information 38 Playback History 40 Metadata 50, 52 Module 54 Track Scoring Module 56 Metadata Maintenance Module 58 Playback History Maintenance Module 60 Metadata Source

Claims (68)

Enabling a user of a first electronic device to control the performance of a multimedia work by the first electronic device;
Obtaining data on the first electronic device from which the user's preferences regarding multimedia works can be inferred,
Another electronic device generates preference information based on the acquired data, wherein the preference includes a negative preference;
A method involving that.   The method of claim 1, wherein the preferences include only negative preferences.   The method of claim 1, wherein the data is obtained based on the user control of the performance of the work. Allowing the user to control the performance of a multimedia work with one or more other electronic devices;
Obtaining, at the one or more other electronic devices, data from which the user's preferences for a multimedia work can be inferred, wherein the preferences include negative preferences;
Generating preference information based on the acquired data from the first electronic device and the one or more other electronic devices, wherein the generated preference information is associated with the user;
The method of claim 1, further comprising:   The method of claim 1, wherein the first electronic device comprises a portable device.   5. A method according to claim 1 or 4, further comprising designating a sequence of performance of a multimedia work for the user based on the preference information.   The method of claim 1, further comprising selecting a multimedia work for distribution to the first electronic device based on the preference information.   The method of claim 4, further comprising selecting a multimedia work for distribution to at least one of the devices based on the preference information.   The method of claim 1, further comprising storing the multimedia work on the first electronic device.   The method of claim 4, further comprising storing the multimedia work on at least one of the electronic devices.   5. The method of claim 1 or 4, further comprising storing a changing body of a multimedia work at the first electronic device, wherein the body of the multimedia work is determined based on the preference information.   The method of claim 1, wherein the data obtained at the first electronic device includes data from which it can be inferred whether a user did not allow the performance of the work to be completed.   The method of claim 4, wherein the data obtained at the one or more other electronic devices includes data from which a user can infer whether or not the performance of the work was not permitted.   The method of claim 12, wherein the data acquired at the first electronic device includes a time when the performance of the work was last completed.   14. The method of claim 13, wherein the data acquired at the one or more other electronic devices includes a time when the performance of the work was last completed.   The method according to claim 1, wherein the user is inferred to hate the work if the performance of the work is not permitted to complete.   The method of claim 1, wherein the multimedia works stored on the first electronic device comprise a larger library subset of multimedia works available to the user.   The method of claim 1, wherein the acquired data is provided from the first electronic device to the other electronic device using an electronic communication channel or a portable storage device.   The method of claim 4, wherein the preferences include only negative preferences.   The method according to claim 1 or 4, wherein the user is verified for identity and the preference information is associated with the user. Receiving preference information derived from a user during the performance of the work on the first electronic device at an online sales site;
Based on the preference information, select a work to recommend to the user for purchase.
The method of claim 1, further comprising:   The method of claim 4, wherein at least one of the one or more other electronic devices comprises a portable electronic device.   23. A method according to claim 5 or 22, wherein the portable electronic device comprises a compressed music player. Maintain each stored library of multimedia works on a library electronic device,
The body of preference information indicating the user's preferences regarding multimedia works is stored on a portable electronic device,
Enabling a user to control the performance of a multimedia work from any of the library electronic devices to the preference information stored on the portable electronic device;
A method involving that. Store a library of multimedia works in a storage device,
Selecting a portion of the multimedia work from the library based on probabilistic user preference information for performance to a user at a location associated with the storage device;
Assessing a fee for playing the selected multimedia work on a performance device.   26. The method of claim 25, wherein the selection of a portion of the multimedia work is based on a preference indicated by the user regarding the performance of the multimedia work.   26. The method of claim 25, wherein the fee is based on the number of works performed.   26. The method of claim 25, wherein the fee includes a one-time fee for all performances of works in the library.   26. The method of claim 25, wherein the fee includes a fee for each use.   26. The method of claim 25, wherein the work is locked for access and the fee is assessed for unlocking access.   26. The method of claim 25, wherein the library is automatically updated on the storage device.   32. The method of claim 31, wherein the update includes adding and / or removing works in the library.   26. The method of claim 25, further comprising paying royalty to a rights holder of the work based on information about the selected multimedia work.   26. The method of claim 25, wherein the location includes a location associated with personal movement.   35. The method of claim 34, wherein the location comprises one of a car, an airport, or a hotel. Preloading a library of multimedia works into a vehicle storage device, and the multimedia works in the library are inaccessible except when unlocked;
Selectively unlocking a portion of the library based on preference information indicating preferences for multimedia works of the user of the vehicle;
A method involving that.   37. The method of claim 36, wherein at least a portion of the preference information is provided by the user during use of the vehicle. Playing the first multimedia work for the user,
Allowing the user to indicate preferences regarding multimedia works similar to the first work to be played later;
A method involving that.   The method of claim 1 or 2, further comprising allowing the user to indicate preferences for multimedia works similar to the first work to be played later. Perform the first multimedia work for the user,
Allowing the user to indicate preferences regarding multimedia works that are not similar to the first work to be played later;
A method involving that.   The method of claim 1 or 2, further comprising allowing the user to indicate preferences for multimedia works similar to the first work to be played later. Manages the body of preference information indicating user preferences for multimedia works that are part of the first library of multimedia works,
Applying the preference information to generate a performance sequence for a second library of multimedia works separate from the first library;
A method involving that.   43. The method of claim 42, wherein the first library and the second library share at least some common works.   43. The method of claim 42, wherein the first library of multimedia works and the second library of multimedia works are available at different locations.   43. The method of claim 42, wherein the preference information is stored on a portable medium and carried from one location to another for use with different libraries of multimedia works.   43. The method of claim 42, wherein the preference information is stored on a network. Manage the first body of preference data or preference information indicating the preferences of the first user for multimedia works,
Manage the second body of preference data or preference information that indicates the preferences of the second user regarding the multimedia work,
Deriving preference data or preference information third body from the first body and the second body;
A method involving that.   48. The method of claim 47, wherein the third body of preference information is used to generate a performance sequence for additional users.   49. The method of claim 48, wherein the derivation includes aggregating the first body and the second body of data or information and processing the aggregation to derive the third body of data or information. .   50. The method of claim 49, wherein the first user and the second user include the same user. Store on the device a set of multimedia works, including a sub-library of a larger library of available multimedia works,
Playing at least some of the multimedia works on the device;
Receiving data indicating user preferences regarding the played multimedia work;
Updating the multimedia work included in the stored set based on the user's preferences;
A method involving that.   52. The method of claim 51, wherein the device comprises a handheld device or an in-vehicle device.   52. The method of claim 51, comprising repeatedly receiving information indicative of a user's preferences and repeatedly updating the multimedia work based on the preferences. Maintain information indicating user preferences regarding played multimedia works,
Probabilistically generating a set of multimedia works based on the information and available multimedia works,
Distribute the set of multimedia works to portable devices;
A method involving that.   55. The method of claim 54, wherein the set of multimedia works is distributed in a compressed format.   55. The method of claim 54, wherein the compression format includes MP3.   55. The method of claim 54, wherein the distribution is performed by recording to disk. The distribution is
55. The method of claim 54, comprising downloading the multimedia work to a portable device.   55. The method of claim 54, wherein the download is performed without the user being aware that the download is taking place.   55. The method of claim 54, wherein the portable device comprises a handheld device or an in-vehicle device.   60. The method of claim 59, wherein the downloading is performed wirelessly. Maintain a body of preference information that shows the user's preferences for multimedia works,
In the performance device, receive preference information or work selection information from the user,
Based on the preference information, the work selection information, or preference information stored in advance regarding the user's preferences regarding the multimedia work, a multimedia work is selected for performance, and one of the selected works is selected. Is not stored in the performance device,
Obtaining the unstored work from an available source and loading it into the playing device without necessarily indicating to the user that a work is not stored on the playing device;
A method involving that. Maintain the multimedia work that is played on a portable device with limited storage,
From time to time, the portable device automatically replaces a work that is less likely to be preferred by a user with a work that is more likely to be preferred.
A method involving that. A device including storage for storing a set of multimedia works, including a larger library subset of available multimedia works;
An element configured to play at least some of the multimedia works on the device;
Stored information indicating user preferences regarding the multimedia work to be performed;
A port for receiving updates of the multimedia work included in the stored set based on the preferences of the user;
With a device. Receive online preference information derived from the user while playing the work on the local device,
Based on the preference information, select works proposed to the user for purchase.
A method involving that. Receiving preference information derived from a user while playing a work on the one or more other electronic devices at an online sales site;
Based on the preference information, select works proposed to the user for purchase.
The method of claim 4 comprising:   An online sales site receives user preference information, selects a work proposed to the user for purchase based on the preference information, and downloads a work selected by the user in response to the proposal 26. The method of claim 25, further comprising assessing a fee for playing the downloaded work. Store a library of multimedia works on a storage device,
Select a portion of the multimedia work from the library for performance to the user;
Assess the fees for playing the selected multimedia work on the performance device;
Preventing access by the user to data including the multimedia work stored in the storage device;
A method involving that.

Images