US20090049045A1

US20090049045A1 - Method and system for sorting media items in a playlist on a media device

Info

Publication number: US20090049045A1
Application number: US11/757,219
Authority: US
Inventors: Waymen J. Askey; Hugh Svendsen
Original assignee: Concert Technology Corp
Current assignee: Concert Technology Corp
Priority date: 2007-06-01
Filing date: 2007-06-01
Publication date: 2009-02-19
Also published as: US20120042245A1; US8983950B2

Abstract

A computer-implemented method and system are provided for sorting media items in a playlist on a media device that are ranked by score. One embodiment includes providing each of the media items with a respective profile score that is based on user preferences and a respective replay score that affects replay of the corresponding media item; and automatically sorting the media items by the replay scores. Another embodiment includes sorting the media items by a sort criteria indicated by a user; sorting the media items by at least one score associated with each of the media items; displaying the playlist; and playing the media items in the playlist according to sort order and automatically skipping media items that have already been played. Yet a further embodiment includes changing a user preference associated with a sort criteria selected by the user, wherein each of the media items includes a profile score that is calculated based on user preferences, and a replay score that affects replay of the corresponding media item; recalculating the profile score and the replay score of each of the media items; sorting the media items by the replay scores; and sorting the media items by the sort criteria indicated by a user.

Description

BACKGROUND OF THE INVENTION

In recent years, there has been an enormous increase in the amount of digital media, such as music, available online. Services such as Apple's iTunes enable users to legally purchase and download music. Other services such as Yahoo! Music Unlimited and RealNetwork's Rhapsody provide access to millions of songs for a monthly subscription fee. As a result, music has become much more accessible to listeners worldwide. However, the increased accessibility of music has only heightened a long-standing problem for the music industry, which is namely the issue of linking audiophiles with new music that matches their listening preferences.
Many companies, technologies, and approaches have emerged to address this issue of music recommendation. Some companies have taken an analytical approach. They review various attributes of a song, such as melody, harmony, lyrics, orchestration, vocal character, and the like, and assign a rating to each attribute. The ratings for each attribute are then assembled to create a holistic classification for the song that is then used by a recommendation engine. The recommendation engine typically requires that the user first identify a song that he or she likes. The recommendation engine then suggests other songs with similar attributions. Companies using this type of approach include Pandora, SoundFlavor, MusicIP, and MongoMusic (purchased by Microsoft in 2000).
Other companies take a communal approach. They make recommendations based on the collective wisdom of a group of users with similar musical tastes. These solutions first profile the listening habits of a particular user and then search similar profiles of other users to determine recommendations. Profiles are generally created in a variety of ways such as looking at a user's complete collection, the playcounts of their songs, their favorite playlists, and the like. Companies using this technology include Last.fm, Music Strands, WebJay, Mercora, betterPropaganda, Loomia, eMusic, musicmatch, genielab, upto11, Napster, and iTunes with its celebrity playlists.
The problem with these traditional recommendation systems is that they fail to consider peer influences. For example, the music that a particular teenager listens to may be highly influenced by the music listened to by a group of the teenager's peers, such as his or her friends. As such, there is a need for a music recommendation system and method that recommends music to a user based on the listening habits of a peer group.

BRIEF SUMMARY OF THE INVENTION

The present invention provides a computer-implemented method and system for sorting media items in a playlist on a media device that are ranked by score. One embodiment includes providing each of the media items with a respective profile score that is based on user preferences and a respective replay score that affects replay of the corresponding media item; and automatically sorting the media items by the replay scores. Another embodiment includes sorting the media items by a sort criteria indicated by a user; sorting the media items by at least one score associated with each of the media items; displaying the playlist; and playing the media items in the playlist according to sort order and automatically skipping media items that have already been played. Yet a further embodiment includes changing a user preference associated with a sort criteria selected by the user, wherein each of the media items includes a profile score that is calculated based on user preferences, and a replay score that affects replay of the corresponding media item; recalculating the profile score and the replay score of each of the media items; sorting the media items by the replay scores; and sorting the media items by the sort criteria indicated by a user.
According to the method and system disclosed herein, the system can serve as a media discovery device by accepting media recommendations from a user's peers and ranking those recommendations for playback by replay score, but allows the user to indicate sort criteria other than replay score. The media items are then sorted for playback based on a combination of both the replay scores and the indicated sort criteria.

BRIEF DESCRIPTION OF SEVERAL VIEWS OF THE DRAWINGS

FIG. 1 illustrates a system incorporating a peer-to-peer (P2P) network for real time media recommendations according to one embodiment.

FIG. 2 is a flow diagram illustrating the operation of the peer devices of FIG. 1 according to one embodiment.

FIG. 3 illustrates the system 10′ according to a second embodiment of the present invention.

FIG. 4 illustrates the operation of the system of FIG. 3 according to one embodiment.

FIG. 5 is a flow diagram illustrating a method for automatically selecting media items to play based on recommendations from peer devices and user preferences according to one embodiment of the present invention.

FIG. 6 illustrates an exemplary graphical user interface (GUI) for displaying a playlist for the peer devices including both local and recommended media items according to an exemplary embodiment.

FIG. 7 is a flow diagram illustrating a process for scoring and controlling the replay of recommended media items using a no repeat factor according to an exemplary embodiment.

FIG. 8 is a diagram illustrating the GUI displaying the playlist after the profile score is updated with the replay score.

FIG. 9 is a flow diagram illustrating a process for visually indicating a replay status of media items on a media device.

FIG. 10 is a diagram illustrating one embodiment for displaying the profile score and the replay score in a GUI using a graphical representation.

FIG. 11 is a diagram illustrating another embodiment for displaying the profile score and the replay score using a graphical representation.

FIG. 12 is a diagram of the GUI displaying a playlist that has been sorted by a category other than score according to one embodiment.

FIG. 13 is a diagram of the GUI displaying a playlist that has been sorted based on recalculated scores and then by the category User according to one embodiment.

FIG. 14 is a block diagram of an exemplary embodiment of the peer device.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to a method and system for sorting media items in a playlist on a media device that are ranked by score. The following description is presented to enable one of ordinary skill in the art to make and use the invention and is provided in the context of a patent application and its requirements. Various modifications to the embodiments and the generic principles and features described herein will be readily apparent to those skilled in the art. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features described herein.
The present invention is mainly described in terms of particular systems provided in particular implementations. However, one of ordinary skill in the art will readily recognize that this method and system will operate effectively in other implementations. For example, the systems, devices, and networks usable with the present invention can take a number of different forms. The present invention will also be described in the context of particular methods having certain steps. However, the method and system operate effectively for other methods having different and/or additional steps not inconsistent with the present invention.
FIG. 1 illustrates a system 10 incorporating a P2P network for providing real time media recommendations according to one embodiment of the present invention. Note that while the exemplary embodiments focus on song recommendations for clarity and ease of discussion, the present invention is equally applicable to providing recommendations for other types of media items, such as video presentations and slideshows. Exemplary video presentations are movies, television programs, and the like. In general, the system 10 includes a number of peer devices 12-16 that are capable of presenting or playing the media items and which are optionally connected to a subscription music service 18 via a network 20, such as, but not limited to, the Internet. Note that while three peer devices 12-16 are illustrated, the present invention may be used with any number of two or more peer devices.
In this embodiment, the peer devices 12-16 are preferably portable devices such as, but not limited to, portable audio players, mobile telephones, Personal Digital Assistants (PDAs), or the like having audio playback capabilities. However, the peer devices 12-16 may alternatively be stationary devices such as a personal computer or the like.
The peer devices 12-16 include local wireless communication interfaces (FIG. 14) communicatively coupling the peer devices 12-16 to form a peer-to-peer (P2P) network. The wireless communication interfaces may provide wireless communication according to, for example, one of the suite of IEEE 802.11 standards, the Bluetooth standard, or the like. Because the peer devices 12-16 are capable of presenting or playing media items whether or not coupled to the P2P network, the peer devices 12-16 may be considered simply as media devices.
The peer device 12 may include a music player 22, a recommendation engine 24, and a music collection 26. The music player 22 may be implemented in software, hardware, or a combination of hardware and software. In general, the music player 22 operates to play songs from the music collection 26. The recommendation engine 24 may be implemented in software, hardware, or a combination of hardware and software. The recommendation engine 24 may alternatively be incorporated into the music player 22. The music collection 26 includes any number of song files stored in one or more digital storage units such as, for example, one or more hard-disc drives, one or more memory cards, internal Random-Access Memory (RAM), one or more associated external digital storage devices, or the like.
In operation, each time a song is played by the music player 22, the recommendation engine 24 operates to provide a recommendation identifying the song to the other peer devices 14, 16 via the P2P network. The recommendation does not include the song. In one embodiment, the recommendation may be a recommendation file including information identifying the song. In addition, as discussed below in detail, the recommendation engine 24 operates to programmatically, or automatically, select a next song to be played by the music player 22 based on the recommendations received from the other peer devices 14, 16 identifying songs recently played by the other peer devices 14,16 and user preferences associated with the user of the peer device 12.
Like the peer device 12, the peer device 14 includes a music player 28, a recommendation engine 30, and a music collection 32, and the peer device 16 includes a music player 34, a recommendation engine 36, and a music collection 38.
The subscription music service 18 may be a service hosted by a server connected to the network 20. Exemplary subscription based music services that may be modified to operate according to the present invention are Yahoo! Music Unlimited digital music service and RealNetwork's Rhapsody digital music service.
FIG. 2 is a flow diagram illustrating operation of the peer device 12 according to one embodiment of the present invention. However, the following discussion is equally applicable to the other peer devices 14,16. First, the peer devices 12-16 cooperate to establish a P2P network (step 200). The P2P network may be initiated using, for example, an electronic or verbal invitation. Invitations may be desirable when the user wishes to establish the P2P network with a particular group of other users, such as his or her friends. Note that this may be beneficial when the user desires that the music he or she listens to be influenced only by the songs listened to by, for example, the user's friends. Invitations may also be desirable when the number of peer devices within a local wireless coverage area of the peer device 12 is large. As another example, the peer device 12 may maintain a “buddy list” identifying friends of the user of the peer device 12, where the peer device 12 may automatically establish a P2P network with the peer devices of the users identified by the “buddy list” when the peer devices are within a local wireless coverage area of the peer device 12.
Alternatively, the peer device 12 may establish an ad-hoc P2P network with the other peer devices 14,16 by detecting the other peer devices 14,16 within the local wireless coverage area of the peer device 12 and automatically establishing the P2P network with at least a subset of the detected peer devices 14, 16. In order to control the number of peer devices within the ad-hoc P2P network, the peer device 12 may compare user profiles of the users of the other peer devices 14,16 with a user profile of the user of the peer device 12 and determine whether to permit the other peer devices 14,16 to enter the P2P network based on the similarities of the user profiles.
At some point after the P2P network is established, the peer device 12 plays a song (step 202). Initially, before any recommendations have been received from the other peer devices 14, 16, the song may be a song from the music collection 26 selected by the user of the peer device 12. Prior to, during, or after playback of the song, the recommendation engine 24 sends a recommendation identifying the song to the other peer devices 14,16 (step 204). The recommendation may include, but is not limited to, information identifying the song such as a Globally Unique Identifier (GUID) for the song, title of the song, or the like; a Uniform Resource Locator (URL) enabling other peer devices to obtain the song such as a URL enabling download or streaming of the song from the subscription music service 18 or a URL enabling purchase and download of the song from an e-commerce service; a URL enabling download or streaming of a preview of the song from the subscription music service 18 or a similar e-commerce service; metadata describing the song such as ID3 tags including, for example, genre, the title of the song, the artist of the song, the album on which the song can be found, the date of release of the song or album, the lyrics, and the like.
The recommendation may also include a list of recommenders including information identifying each user having previously recommended the song and a timestamp for each recommendation. For example, if the song was originally played at the peer device 14 and then played at the peer device 16 in response to a recommendation from the peer device 14, the list of recommenders may include information identifying the user of the peer device 14 or the peer device 14 and a timestamp identifying a time at which the song was played or recommended by the peer device 14, and information identifying the user of the peer device 16 or the peer device 16 and a timestamp identifying a time at which the song was played or recommended by the peer device 16. Likewise, if the peer device 12 then selects the song for playback, information identifying the user of the peer device 12 or the peer device 12 and a corresponding timestamp may be appended to the list of recommenders.
The peer device 12, and more specifically the recommendation engine 24, also receives recommendations from the other peer devices 14, 16 (step 206). The recommendations from the other peer devices 14,16 identify songs played by the other peer devices 14, 16. Optionally, the recommendation engine 24 may filter the recommendations from the other peer devices 14, 16 based on, for example, user, genre, artist, title, album, lyrics, date of release, or the like (step 208).
The recommendation engine 24 then automatically selects a next song to play from the songs identified by the recommendations received from the other peer devices 14, 16, optionally songs identified by previously received recommendations, and one or more songs from the music collection 26 based on user preferences (step 210). In one embodiment, the recommendation engine 24 considers only those songs identified by recommendations received since a previous song selection. For example, if the song played in step 202 was a song selected by the recommendation engine 24 based on prior recommendations from the peer devices 14, 16, the recommendation engine 24 may only consider the songs identified in new recommendations received after the song was selected for playback in step 202 and may not consider the songs identified in the prior recommendations. This may be beneficial if the complexity of the recommendation engine 24 is desired to be minimal such as when the peer device 12 is a mobile terminal or the like having limited processing and memory capabilities. In another embodiment, the recommendation engine 24 may consider all previously received recommendations, where the recommendations may expire after a predetermined or user defined period of time.
As discussed below, the user preferences used to select the next song to play may include a weight or priority assigned to each of a number of categories such as user, genre, decade of release, and location/availability. Generally, location identifies whether songs are stored locally in the music collection 26; available via the subscription music service 18; available for download, and optionally purchase, from an e-commerce service or one of the other peer devices 14, 16; or are not currently available where the user may search for the songs if desired. The user preferences may be stored locally at the peer device 12 or obtained from a central server via the network 20. If the peer device 12 is a portable device, the user preferences may be configured on an associated user system, such as a personal computer, and transferred to the peer device 12 during a synchronization process. The user preferences may alternatively be automatically provided or suggested by the recommendation engine 24 based on a play history of the peer device 12. In the preferred embodiment discussed below, the songs identified by the recommendations from the other peer devices 14, 16 and the songs from the music collection 26 are scored or ranked based on the user preferences. Then, based on the scores, the recommendation engine 24 selects the next song to play.
Once the next song to play is selected, the peer device 12 obtains the selected song (step 212). If the selected song is part of the music collection 26, the peer device 12 obtains the selected song from the music collection 26. If the selected song is not part of the music collection 26, the recommendation engine 24 obtains the selected song from the subscription music service 18, an e-commerce service, or one of the other peer devices 14, 16. For example, the recommendation for the song may include a URL providing a link to a source from which the song may be obtained, and the peer device 12 may obtain the selected song from the source identified in the recommendation for the song. Once obtained, the selected song is played and the process repeats (steps 202-212).
FIG. 3 illustrates the system 10′ according to a second embodiment of the present invention. In this embodiment, the peer devices 12′-16′ form a P2P network via the network 20 and a proxy server 40. The peer devices 12′-16′ may be any device having a connection to the network 20 and audio playback capabilities. For example, the peer devices 12′-16′ may be personal computers, laptop computers, mobile telephones, portable audio players, PDAs, or the like having either a wired or wireless connection to the network 20. As discussed above with respect to the peer device 12, the peer device 12′ includes music player 22′, a recommendation engine 24′, and a music collection 26′. Likewise, the peer device 14′ includes a music player 28′, a recommendation engine 30′, and a music collection 32′, and the peer device 16′ includes a music player 34′, a recommendation engine 36′, and a music collection 38.
FIG. 4 is a flow diagram illustrating operation of the system 10′ as shown in FIG. 3. In this example, once the P2P network is established, the peer device 14′ plays a song and, in response, provides a song recommendation identifying the song to the peer device 12′ via the proxy server 40 (steps 400-404). While not illustrated for clarity, the peer device 14′ also sends the recommendation for the song to the peer device 16′ via the proxy server 40. The peer device 16′ also plays a song and sends a song recommendation to the peer device 12′ via the proxy server 40 (steps 406-410). Again, while not illustrated for clarity, the peer device 16′ also sends the recommendation for the song to the peer device 14′ via the proxy server 40. From this point, the process continues as discussed above.
FIG. 5 illustrates a process of automatically selecting a song to play from the received recommendations and locally stored songs at the peer device 12′ according to one embodiment of the present invention. However, the following discussion is equally applicable to the peer devices 12-16 of FIG. 1, as well as the other peer devices 14′-16′ of FIG. 3. First, the user preferences for the user of the peer device 12′ are obtained (step 500). The user preferences may include a weight or priority assigned to each of a number of categories such as, but not limited to, user, genre, decade of release, and location/availability. The user preferences may be obtained from the user during an initial configuration of the recommendation engine 24′. In addition, the user preferences may be updated by the user as desired. The user preferences may alternatively be suggested by the recommendation engine 24′ or the proxy server 40 based on a play history of the peer device 12′. Note that proxy server 40 may ascertain the play history of the peer device 12′ by monitoring the recommendations from the peer device 12′ as the recommendations pass through the proxy server 40 on their way to the other peer devices 14′-16′. The user preferences may be stored locally at the peer device 12′ or obtained from a central server, such as the proxy server 40, via the network 20.
Once recommendations are received from the other peer devices 14′-16′, the recommendation engine 24′ of the peer device 12′ scores the songs identified by the recommendations based on the user preferences (step 502). The recommendation engine 24′ also scores one or more local songs from the music collection 26′ (step 504). The recommendation engine 24′ then selects the next song to play based, at least in part, on the scores of the recommended and local songs (step 506).
FIG. 6 illustrates an exemplary graphical user interface (GUI) 42 for displaying a playlist for the peer devices including both local and recommended media items according to an exemplary embodiment. In this example, the media items displayed in the playlist are songs, and information for each song is displayed in several category fields of the playlist. In this embodiment, the categories are users, genre, decade, and location/availability, but it may include other types of categories. In this example, the peer device 12′ plays media items from a playlist that includes a mixture of items selected by the user of the device (in this case Hugh) and recommended media items from the user's friends (in this case Gary, Gene, Mike, and Waymen). The playlist is continually updated as recommendations are received. Note that the playlist shows a mixture of the media items that are on the user's machine (designated by a location Local) and items that have been recommended from friends (Gary, Gene, Mike, and Waymen) that may need to be downloaded, or can be streamed from within the music subscription service 18.
In this example, both the local and recommended songs are scored based on the category weights, and sorted according to their scores. The weights for the categories may be assigned manually by the user via a GUI of the peer device 12 or a website (e.g., subscription music service 18), or assigned based on a user profile. In an exemplary embodiment, the peer device 12′ always plays the item with the highest score, which in this embodiment is the song at the top of the playlist.
Media items can be scored a number of different ways utilizing various mechanisms and formulas. According to an exemplary embodiment, one equation for scoring the media items as a function of the weighted categories (and subcategories) is:
Score=( 1/10)*(1/(WD+WG+WL+WU))*(WD*WDA+WG*WGA+WL*WLA+WU*WUA)*100
where WU is the weight assigned to the user category; WUA is the weight assigned to the user attribute of the song, which is the user recommending the song (e.g., Hugh, Gary, Gene, et al); WG is the weight assigned to the genre category; WGA is the weight assigned to the genre attribute of the song, which is the genre of the song (e.g., Alternative, Rock, Jazz, Punk, etc.); WD is the weight assigned to the decade category; WDA is the weight assigned to the decade attribute of the song, which is the decade in which the song or the album associated with the song was released (e.g., 1960, 1970, etc.); WL is the weight assigned to the location/availability category; and WLA is the weight assigned to the location/availability attribute of the song, which is the location or availability of the song (e.g., Local, Subscription, Download, etc.).
As an example, assume that the following weights have been assigned to the categories as follows:
User Category 1

Genre Category 7

Decade Category 7

Location/Availability Category 5

Further assume that attributes for the categories have been assigned weights as follows:


User	Genre	Decade	Location/Availability

Hugh

	9	Alternative	8	1950s	2	Local	8
Gary	5	Classic Rock	5	1960s	4	Subscription Network	2
Gene	5	Arena Rock	5	1970s	7	Buy/Download	1
		Jazz	5	1980s	9	Find	1
		New Wave	2	1990s	5
		Punk	4	2000s	5
		Dance	2
		Country	2

Inserting these weights into the score equation for the song “Say Hey” in FIG. 6 yields:

Score=( 1/10)*(1/(7+7+5+1))*(7*9+7*8+5*8+1*9)*100
Score=( 1/10)*( 1/20)*(63+56+40+9)*100
Score=( 1/10)*( 1/20)*(168)*100
Score=84
In the playlist shown in FIG. 6, note that the song “Say Hey” by “The Tubes” is the first item played in the playlist because it has the highest score according to the category weights described above. In one embodiment the score is calculated based on the category weights in the user's preferences or profile. The score is referred to hereinafter as a profile score. However, those with ordinary skill in the art will readily recognize that the profile score may be based on other factors other than category weights.
Scoring and Affecting the Replay of Recommended Media Items Using a No Repeat Factor
It would be undesirable to most users if any particular media item is repeatedly replayed within a short time interval. However, if the peer device 12′ plays the media item with the highest profile score and the user does not receive any new recommendations with a higher profile score than the media item already played, then the peer device 12′ could repeatedly play the same media item, absent a mechanism for altering replay of media items.
According to a further aspect of the invention, in response to each one of the media items being played, the peer device 12′ calculates a respective replay score for the media item that affects or influences replay of the media item. In one embodiment, the replay score is calculated at least in part as a function of a no repeat factor (NRF). The replay scores of the media items can then be used to sort the media items for playing.
In one embodiment, the NRF is based on a user settable value. For example, a weighted no repeat (WNR) category may be assigned a value of 9 out of 10, meaning that the period between repeated playings should be longer rather than shorter. In another embodiment, the NRF may be based on the total number of media items in the playlist, rather than a fixed WNR.
FIG. 7 is a flow diagram illustrating a process for scoring and affecting the replay of recommended media items using a no repeat factor according to one embodiment. The process assumes that the recommendation engine 24′ has already calculated the profile scores of each of the media items in the playlist. The process begins in response to one of the media items being played (step 700), i.e., the song at the top of the playlist, at which time the recommendation engine 24′ calculates a no repeat factor (NRF) for the media item as a function of the weighted no repeat (WNR) value (step 702).
In one embodiment the NRF may be calculated using the formula:
$NRF = \frac{MIN (10 \cdot WNR, LASTREPEAT_INDEX)}{10 \cdot WNR}$
where the LastRepeat_Index is preferably based on one or both of a count of the number of media items played since the last play of the media item, or a predetermined time period, e.g., 2 hrs, 5 hrs, 1 day, and so on.
For example, referring to the playlist shown in FIG. 6, after “Say Hey” has been played, the number of songs since this song was last played is now 1. Assuming the weighted no repeat value (WNR) is 9, the NRF can be computed as follows:
No Repeat Factor=Min[10*WNR, LastRepeat_Index]/(10*WNR)
No Repeat Factor=Min [10*9,1]/(10*9)
No Repeat Factor=1/(10*9)
No Repeat Factor=0.0111
In this embodiment, it should be understood that the weighted no repeat (WNR) value may be a global variable that applies equally to each of the user's media items, while the last repeat index and the corresponding no repeat factor (NRF) may be different for each of the media items. Each time a media item is played, the last repeat index is incremented/decremented or calculated for each of the media items that have already been played. For example, if the last repeat index is based on the number of songs played since the last play of the media item, then the last repeat index is incremented. If the last repeat index is based on a predetermined time period, then the last repeat index could be calculated to determine how much time has passed since the last play of the media item, e.g., based on the difference between the time the last play occurred and the current time.
As stated above, in one embodiment the NRF may be based on the number of media items in the playlist, which is dynamic. In this embodiment, the WNR can be replaced by the total number of media items in the playlist, which ensures that each item will not be repeated based in part until most or all of the other items have been played. Thus, the NRF scales naturally to the size of the playlist.
Next, the recommendation engine 24′ calculates a replay score for the media item (as well as for the other previously played media items) based on a function of the category weights and the NRF (step 704). In one embodiment, the replay score may be computed using the equation:
Replay Score=NRF*( 1/10)*(1/(WD+WG+WL+WU))*(WD*WDA+WG*WGA+WL*WLA+WU*WUA)*100
Continuing with the example playlist shown in FIG. 6, the replay score for the song “Say Hey” immediately after it was played (or while it was playing) and after computation of the NRF would be:
Replay Score=(0.011)*( 1/10)*(1/(7+7+5+1))*(7*9+7*8+5*8+1*9)*100
Replay Score=(0.011)*( 1/10)*( 1/20)*(63+56+40+9)*100
Replay Score=(0.011)*( 1/10)*( 1/20)*(168)*100
Replay Score=0.9
Replay Score ˜=1
Referring again to FIG. 7, in one embodiment, the recommendation engine 24′ updates the profile scores of media items with the corresponding replay scores, and re-orders the playlist based on the updated profile scores (step 706).
FIG. 8 is a diagram illustrating the GUI 42 displaying the playlist after the profile score for the song “Say Hey” is updated with the replay score. Since the song “Say Hey” has a replay score of 1, and the replay score is used to update the profile score, the profile score becomes 1, and the song “Say Hey” drops to the bottom of the playlist, ensuring that the song will not be repeated before other songs have a chance to play.
The first aspect of the exemplary embodiment provides a P2P network for real-time media recommendations in which peer devices constantly receive recommendations of media items from other peer devices; intersperses the recommendations with an existing playlist of media items designated by a user; dynamically calculates both a profile score of each of the media items according to the user's preferences, and a replay score for previously played media items that affects replay of the media items; and uses the replay score to update the profile score in order to play the media items back in score order. This embodiment ensures that there are no repetitions of played media items until the user has had at least some exposure to other recommended media items in the playlist.
Visually Indicating a Replay Status of a Media Item
While the replay score ensures that the user will have some exposure to other media items in the playlist before repeating the media items that have already been played, the replay score can sometimes have the effect of hiding the media items that the user most likely will enjoy by placing those items at the end of the playlist. Continuing with the example given above, for instance, the song “Say Hey” had an original profile score of 84 and was the highest in the playlist. This means that “Say Hey” was most likely a song that the user (Hugh) was going to enjoy from the list, given the category weights that the user entered in the system (this assumes that the user has set the weights in the system to yield songs that most closely match his tastes). Once the song has been played, though, the replay score is calculated, and the song “Say Hey” has a score of 1. The user might forget that this song was once at the top of the list, given its current score.
Accordingly, a further aspect of the present invention provides a mechanism for visually indicating the replay status of a media item by letting the user see the original profile score of the media item as well as the current replay score, as determined, for example, by the no repeat factor. In this embodiment, the peer devices 12-16 retain the two scores and provide a GUI to clearly indicate both pieces of information to the user. By displaying the replay score, the user is apprised of the replay status of one or all of the media items.
FIG. 9 is a flow diagram illustrating a process for visually indicating a replay status of media items on a media device. The process begins by displaying the media items in a graphical user interface (GUI) of the media device (step 900). As shown in FIGS. 6 and 8, in the exemplary embodiment, the GUI displays the media items in a playlist, which are represented by text information, such as song title. However, the media items could also be displayed with graphical representations, such as icons and/or pictures (e.g., album covers).
As also described above, the profile scores of the media items that are calculated based on user preferences are also displayed in the GUI 42 (step 902). However, according to this embodiment, the GUI 42 can also display the replay scores for the media items that affect the replay of the media items (step 904). As described above, the replay scores can be based on corresponding no repeat factors (NRF), which in turn, can be derived from either a predetermined time period and/or a count of media items that have been played since the first media item was last played.
The media items are sorted in the playlist based on the replay scores (step 906). In one embodiment, all media items in the playlist are provided with replay scores whether or not the media item has been played, with the initial values for replay scores being set equal to the profile score of the corresponding media item. In another embodiment, all the media items have a profile score, but replay scores are only calculated after the corresponding media items have been played. In this case, the sorting can be controlled by the replay scores for previously played media items that have respective replay scores, and by the profile score for the media items that have not yet been played on the peer device and only have profile scores (step 906). As a practical matter, during operation of the peer device, the sorting of the playlist (step 906) may occur prior to display of the playlist (steps 902-904).
Based on the above, it should become apparent that the profile score is a relatively fixed value that is determined through the interaction of the user's profile/preferences with a given media item. However, the replay score is a dynamic value that will normally range between, but is not limited to, a maximum of the profile score and a lesser value determined by the no repeat factor (NRF).
There are several embodiments for indicating both the profile score and the replay score for each media item. In one embodiment, a representation of the replay score relative to the profile score is displayed in association with the media item.
FIG. 10 is a diagram illustrating one embodiment for displaying the profile score and the replay score in a GUI using a graphical representation. In this embodiment, the profile score 1000 and replay score 1002 are shown displayed using bar graphs. One bar graph displays the profile score 1000 relative to a maximum profile score 1004 (e.g., max. 100), and a second bar graph displays the replay score 1002 relative to the profile score 1000. In this example, the second bar graph is shown having a length that indicates the profile score 1000 and a shaded subsection that indicates the replay score 1002. In addition, the second bar graph is also shown to display numeric values for both the profile score 1000 and replay score 1002, e.g., “1 of 84”.
FIG. 11 is a diagram illustrating another embodiment for displaying the profile score 1000 and the replay score 1002 using a graphical representation. In this embodiment, the profile score 1000 and the replay score 1002 are displayed in the single bar graph of FIG. 10 that shows both the profile score 1000 and the replay score 1002, which makes it suitable for display in the playlist GUI 42 next to each media item.
Although bar graphs have been described for graphically illustrating the profile score 1000 and the replay score 1002, the profile score 1000 and the replay score 1002 could be displayed using other graphic representations, such as a pie chart. The profile score 1000 and the replay score 1002 may also be displayed with just text information. For example, the replay score 1002 may be displayed as a percentage of the profile score, such as 4.5%, for instance.
Referring again to FIG. 9, the calculation of the NRF is such that the replay score 1002 for a particular media item is allowed to recharge back to the value of the profile score 1000 as media items are played and/or time passes (step 908) with a corresponding display in the display of the scores. For example, as shown in FIG. 11, four songs have played since the first playing of the song “Say Hey”, and the replay score 1002 for the song has increased accordingly from an initial value of 1 to a current value of 5.
Referring again to FIG. 9, once the replay score 1002 of the media item reaches the value of the profile score 1000, only the profile score 1000 for the media item is shown in the GUI (step 910).
Given the above description of the profile and replay scores 1002, it should be apparent that the exemplary embodiments cover alternative embodiments that include a wider range of category weightings and accompanying profile scores 1000 and presentation factors controlling playback beyond the no repeat factor (NRF), such as for example, a methodology that attempts to force play back of a media item based solely on time, such as at least once per week or alternatively, no more often than once per day. Also, although described in terms of a P2P media recommendation environment, the exemplary embodiments may be applied to media devices in traditional client/server environments as well.
Sorting Recommended Media Items in a Scored Playlist
One purpose of the P2P networked media recommendation system 10 is to provide a music discovery mechanism for the user. While one purpose of creating a playlist of recommendations is the creation of a musical journey for the user, it is entirely possible that users of the media recommendation system 10 may want to sort on different categories as a means to quickly peruse the recommendations that have been received from their peers. For example, maybe the user Hugh would like to quickly see how many recommendations have been received by a particular friend (Waymen) and the associated scores 1000 and 1002 of such media items.
FIG. 12 is a diagram of the GUI 42 displaying a playlist that has been sorted by a category other than score according to one embodiment. Sorting in this fashion can pose a potential problem for the media recommendation system 10 since the media recommendation system 10 is designed to play the first media item in the sorted playlist. If the user sorts by some column other than score, then it is possible the peer devices 12-16 will begin playing media items that are either (1) not the media items most likely to match the user's tastes or (2) may be media items that have already been played before (thereby eliminating the music discovery aspects of the application).
According to a further aspect of the exemplary embodiment, embodiments for sorting the playlist are provided that maintain the system's purpose as a media discovery device by accepting media recommendations from a user's peers and by ranking those recommendations for playback by score, but also allows the user to indicate a sort criteria other than score. The media items are then sorted for playback based on a combination of both the score and the indicated sort criteria.
In one embodiment, the peer devices 12-16 permit the sorting of the playlist by different category columns, but only subordinate to a sort by score. In this embodiment, each of the media items include a profile score 1000 and a replay score 1002. First, the peer devices 12-16 automatically sort the media items in the playlist by the replay scores 1002. As stated above, the replay score 1002 may be set equal to the profile score 1000 for the media items that have yet to be played. Second, the peer devices 12-16 sort the media items by a sort criteria indicated by a user. For example, if the user wants to sort on the User column, then the peer devices 12-16 perform a double sort where the media items in the playlist are first sorted by the profile and replay scores and then by User. Finally, the playlist is displayed and the media items in the playlist are played according to the sort order. The steps of sorting and displaying the playlist are not necessarily order dependent.
In a second embodiment, the peer devices 12-16 permit the user to sort the playlist by category columns other than score first and then sort by score second. In this embodiment, the peer devices 12-16 first sort the media items by a sort criteria indicated by a user. For example, the user may select a particular category to sort on by clicking one of the category columns in the playlist. Thereafter, the peer devices 12-16 sort the media items by the score associated with each of the media items, e.g., the profile and replay scores 1000 and 1002, and displays the sorted playlist. To preserve the integrity of the recommendation engine 24 as a music discovery device, the media items in the playlist are played according to sort order, but the media items that have already been played (as indicated by a corresponding replay score 1002), are automatically skipped.
In this particular case, the playlist would look similar to that of FIG. 12, which shows an example playlist that has been sorted by User first, and then sorted by score second. However, in this embodiment as songs are played by moving down the playlist, songs that have already been played are automatically skipped. Notice that in this embodiment, the media item being played is not necessarily the first item in the playlist, as in the case where the first media item is a replay score.
In a third embodiment, the peer devices 12-16 permit the user to sort the playlist by category columns other than score, but adjust the weight of the selected category so that the selected category has a greater weight than the other categories listed by the user. In response to receiving the user's selection of sort criteria, such as selecting a particular category to sort on by clicking one of the category columns in the playlist, a user preference associated with the sort criteria is changed. As described above, user preferences used to select the next song to play may include a weight assigned to each of a number of categories, such as user, genre, decade of release, and location/availability. The category weights are then used to score or rank the media items from the music collection 26.
As an example, suppose the user chooses to sort the playlist by User. Then, in this embodiment, the User weight (WU) may be increased automatically from its initial value of 1 (see FIG. 8) to near a maximum value, such as 9 for instance, thereby making it a dominant force in the user's preferences and calculation of the profile score 1000.
After the user preference associated with the sort criteria is changed, the profile score 1000 and any existing replay score 1002 are recalculated. The media items in the playlist are then first sorted by the recalculated replay scores 1002, as described above, and then sorted by the sort criteria selected by the user, e.g., by the category User. The sorted playlist is displayed and the media items are played in the playlist according to sort order.
FIG. 13 is a diagram of the GUI 42 displaying a playlist that has been sorted based on recalculated scores and then by the User category according to one embodiment. Because of the change in the weight assigned to the user category, the media items now have slightly different profile and replay scores 1000 and 1002.
FIG. 14 is a block diagram of an exemplary embodiment of the peer device 12′ of FIG. 3. However, the following discussion is equally applicable to the other peer devices 14′-16′, as well as peer devices 12-16 of FIG. 1. In general, the peer device 12′ includes a control system 154 having associated memory 156. In this example, the music player 22′ and the recommendation engine 24′ are at least partially implemented in software and stored in the memory 156. The peer device 12′ also includes a storage unit 158 operating to store the music collection 26′ (FIG. 3). The storage unit 158 may be any number of digital storage devices such as, for example, one or more hard-disc drives, one or more memory cards, RAM, one or more external digital storage devices, or the like. The music collection 26′ may alternatively be stored in the memory 156. The peer device 12′ also includes a communication interface 160. The communication interface 160 includes a network interface communicatively coupling the peer device 12′ to the network 20 (FIG. 3). The peer device 12′ also includes a user interface 162, which may include components such as a display, speakers, a user input device, and the like.
The present invention provides substantial opportunity for variation without departing from the spirit or scope of the present invention. For example, while FIG. 1 illustrates the peer devices 12-16 forming the P2P network via local wireless communication and FIG. 3 illustrates the peer devices 12′-16′ forming the P2P network via the network 20, the present invention is not limited to either a local wireless P2P network or a WAN P2P network in the alternative. More specifically, a particular peer device, such as the peer device 12, may form a P2P network with other peer devices using both local wireless communication and the network 20. Thus, for example, the peer device 12 may receive recommendations from both the peer devices 14, 16 (FIG. 1) via local wireless communication and from the peer devices 14′-16′ (FIG. 3) via the network 20.
A method and system for sorting media items in a playlist on a media device that are ranked by score has been disclosed. The present invention has been described in accordance with the embodiments shown, and one of ordinary skill in the art will readily recognize that there could be variations to the embodiments, and any variations would be within the spirit and scope of the present invention. For example, the present invention can be implemented using hardware, software, a computer readable medium containing program instructions, or a combination thereof. Software written according to the present invention is to be either stored in some form of computer-readable medium such as memory or CD-ROM, or is to be transmitted over a network, and is to be executed by a processor. Consequently, a computer-readable medium is intended to include a computer readable signal, which may be, for example, transmitted over a network. Accordingly, many modifications may be made by one of ordinary skill in the art without departing from the spirit and scope of the appended claims.

Claims

1. A computer-implemented method for sorting media items in a playlist, comprising:

providing each of the media items with a respective profile score that is based on user preferences;

providing each of the media items with a respective replay score that affects replay of the corresponding media item;

automatically sorting the media items by the replay scores.

2. The computer-implemented method of claim 1 further comprising: after sorting the media items by the replay scores, sorting the media items by a sort criteria indicated by a user.

3. The computer-implemented method of claim 2 further comprising displaying the playlist and playing the media items in the playlist according to sort order.

4. The computer-implemented method of claim 2 wherein the playlist was created in a networked peer-to-peer media recommendation system.

5. The computer-implemented method of claim 2 further comprising:

in response to each one of the media items being played, recalculating the replay scores for the media items.

6. The computer-implemented method of claim 5 wherein the replay scores recharge over time back to a value of the profile score.

7. The computer-implemented method of claim 2 wherein the sort criteria indicated by the user comprises a selection of at least one category.

8. The computer-implemented method of claim 7 wherein the categories include at least one of a user category, a genre category, a decade category, and a location category.

9. An executable software product stored on a computer-readable medium containing program instructions for sorting media items in a playlist, the program instructions for:

providing each of the media items with a respective replay score that affects replay of the corresponding media item; and

automatically sorting the media items by the replay scores.

10. A computer-implemented method for sorting media items on a media device, comprising:

sorting the media items by a sort criteria indicated by a user;

sorting the media items by at least one score associated with each of the media items;

displaying the playlist; and

playing the media items in the playlist according to sort order and automatically skipping media items that have already been played.

11. The computer-implemented method of claim 10 wherein the playlist was created in a networked peer-to-peer media recommendation system.

12. The computer-implemented method of claim 10 wherein each one of the media items include a profile score that is based on user preferences, and previously played media items include a replay score that affects replay of the corresponding media item.

13. The computer-implemented method of claim 12 further comprising sorting the media items in the playlist based on the profile scores and replay scores, wherein the sorting is controlled by the replay scores.

14. The computer-implemented method of claim 13 further comprising:

in response to each one of the media items being played, recalculating the replay scores.

15. The computer-implemented method of claim 14 wherein the replay scores recharge over time back to a value of the profile score.

16. The computer-implemented method of claim 11 wherein the sort criteria indicated by the user comprises a selection of at least one category.

17. The computer-implemented method of claim 16 wherein the categories include at least one of a user category, a genre category, a decade category, and a location category.

18. An executable software product stored on a computer-readable medium containing program instructions for sorting media items in a playlist, the program instructions for:

sorting the media items by a sort criteria indicated by a user;

displaying the playlist; and

19. A computer-implemented method for sorting media items, comprising:

changing a user preference associated with a sort criteria selected by the user, wherein each of the media items includes a profile score that is calculated based on user preferences, and a replay score that affects replay of the corresponding media item;

recalculating the profile score and the replay score of each of the media items;

sorting the media items by the replay scores; and

sorting the media items by the sort criteria indicated by a user.

20. The computer-implemented method of claim 19 further comprising displaying the playlist and playing the media items in the playlist according to sort order.

21. The computer-implemented method of claim 20 further comprising:

22. The computer-implemented method of claim 21 wherein the replay scores recharge over time back to a value of the profile score.

23. The computer-implemented method of claim 20 wherein the sort criteria indicated by the user comprises a selection of at least one category.

24. The computer-implemented method of claim 23 wherein the categories include at least one of a user category, a genre category, a decade category, and a location category.

25. An executable software product stored on a computer-readable medium containing program instructions for sorting media items in a playlist, the program instructions for:

sorting the media items by the replay scores; and

sorting the media items by the sort criteria indicated by a user.