JP2006293246A - Information processing device, method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To generate a play list such that similarities between two successively reproduced contents vary having certain fluctuations. <P>SOLUTION: Musical pieces to be reproduced are determined while reference is varied on the basis of a reference table. A starting musical piece (musical piece at the head of the play list) is selected by, for example, a user, a 2nd musical piece to be played right after the starting musical piece is a musical piece having a similarity of 0.8 to 1.0 (within a similarity range of reference with a reference number 1) to the starting musical piece, a 3rd musical piece is a musical piece having a similarity of 0.8 to 1.0 (similarity range of reference with a reference number 2) to the 2nd musical piece, and a 4th musical piece is a musical piece having a similarity of 0.6 to 0.8 (similarity range of reference with a reference number 3) to the 3rd musical piece. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an information processing device and method, and a program, and more particularly, to an information processing device and method, and a program for generating a playlist that enables music reproduction so that a user does not get bored.

  2. Description of the Related Art Some playback apparatuses that play back music can generate a list (hereinafter referred to as a playlist) indicating the playback order of a plurality of music and can continuously play back music according to the play list.

  For example, the distance (cosine function) between the user's preference information generated by accumulating the attributes (title, artist, genre, review, etc.) of the music that the user has heard so far and the reproducible music is obtained. A playlist is generated in which music is played continuously.

  In addition, a playlist is generated in which music pieces that are close to the music piece that is played one before are played continuously.

  By the way, when a playlist is generated by such a method, basically, songs having similar tunes are continuously played (for example, the user preference information prefers a song with a slow preference). In the case where it is represented, and when the music played back one time is a slow music, the slow music is played continuously), there is a problem that the user gets bored.

  As a method for solving this problem, a method has been proposed in which a playlist is generated so that specific attributes of music such as tempo of the music to be played change (see Patent Document 1).

JP 2002-333892 A

  However, according to this method, a specific attribute of a music can be changed, but the similarity as a whole music cannot be changed.

  In addition, since the change in the attribute is not determined in relation to the user's preference, the change is not related to the user's preference, and eventually the user's tiredness cannot be avoided.

  The present invention has been made in view of such a situation, and is capable of generating a playlist that enables music reproduction so that the user does not get bored.

  The information processing apparatus according to the present invention continuously reproduces content based on the similarity calculated by the calculating unit, the calculating unit that calculates the similarity between the content whose playback order is determined and other reproducible content. And determining means for determining the reproduction order so that the similarity between the two contents changes with a certain fluctuation.

  Content determining means for determining the content to be reproduced next to the content whose reproduction order has been determined from content whose similarity calculated by the calculating means is a similarity corresponding to a predetermined criterion Changing means for changing the reference can be provided so that the degree of similarity between the two contents to be reproduced changes with a certain fluctuation.

  User preference generation means for generating user preference is further provided, and the determination means can cause content corresponding to the user preference to be the content to be reproduced next in a certain cycle.

  The criterion can be a range of similarities.

  The criterion may be a probability that a content whose similarity with a content whose playback order has been determined is a predetermined similarity or higher will be played next.

  The determination means can make the content selected by the user the content to be reproduced first.

  User preference generation means for generating user preferences can be further provided, and the determination means can determine the content to be reproduced first based on the similarity between the user preference and the reproducible content.

  User preference generation means for generating a user preference is further provided, and the content determination means determines the content to be reproduced next from the content corresponding to the reference based on the similarity calculated by the calculation means. Can be made.

  Separation plane generation means for generating a separation plane that likes and dislikes from user preferences, and the content determination means corresponds to the similarity calculated by the calculation means, and from the contents belonging to the favorite plane of the separation plane, The content to be played next can be determined.

  In the content determination means, among the contents whose similarity calculated by the calculation means is the similarity corresponding to the reference, the content having the musical feature most musically continuous with the content whose reproduction order is determined is The content can be played back.

  The information processing method of the present invention is based on the calculation step for calculating the similarity between the content whose playback order is determined and other content that can be played back, and the similarity calculated in the processing of the calculation step. And determining a content to be reproduced next to the content whose reproduction order has been determined so that the similarity between the two content to be reproduced changes with a certain fluctuation.

  The program of the present invention plays back continuously based on the calculation step for calculating the similarity between content whose playback order is determined and other content that can be played back, and the similarity calculated in the processing of the calculation step. And a determination step of determining a content to be reproduced next to the content whose reproduction order has been determined so that the similarity between the two contents changes with a certain fluctuation.

  In the information processing apparatus, method, and program according to the present invention, the similarity between content whose playback order is determined and other content that can be played back is calculated, and playback is performed continuously based on the calculated similarity. The reproduction order is determined so that the degree of similarity between the two contents to be changed changes with a certain fluctuation.

  ADVANTAGE OF THE INVENTION According to this invention, the play list which enables the music reproduction which a user does not get tired can be produced | generated.

  Embodiments of the present invention will be described below. Correspondences between constituent elements described in the claims and specific examples in the embodiments of the present invention are exemplified as follows. This description is to confirm that specific examples supporting the invention described in the claims are described in the embodiments of the invention. Therefore, even though there are specific examples that are described in the embodiment of the invention but are not described here as corresponding to the configuration requirements, the specific examples are not included in the configuration. It does not mean that it does not correspond to a requirement. On the contrary, even if a specific example is described here as corresponding to a configuration requirement, this means that the specific example does not correspond to a configuration requirement other than the configuration requirement. not.

  Further, this description does not mean that all the inventions corresponding to the specific examples described in the embodiments of the invention are described in the claims. In other words, this description is an invention corresponding to the specific example described in the embodiment of the invention, and the existence of an invention not described in the claims of this application, that is, in the future, a divisional application will be made. Nor does it deny the existence of an invention added by amendment.

An information processing apparatus according to claim 1 is provided.
A calculation means (for example, the similarity calculation unit 24 in FIG. 2) for calculating the similarity between the content whose playback order is determined and other content that can be played back;
Based on the similarity calculated by the calculation means, a determination means (for example, the playback in FIG. 2) that determines the playback order so that the similarity between two pieces of content that are played back continuously changes with a certain fluctuation. And a music determination unit 25).

The information processing apparatus according to claim 2
The decision means is
Content determination means (for example, step S10 in FIG. 4) that determines the content to be reproduced next to the content whose reproduction order has been determined from the content whose similarity calculated by the calculation means is the similarity corresponding to the predetermined reference. 2 to perform the process of step S13, the reproduction music determination unit 25) of FIG.
Changing means for changing the reference so that the similarity between two pieces of content that are played back continuously changes with a certain fluctuation (for example, the playback music shown in FIG. 2 that performs the processing of steps S17 to S19 in FIG. 4) And a determination unit 25).

An information processing apparatus according to claim 3 is provided.
It further comprises user preference generation means (for example, user preference vector generation unit 22 in FIG. 2) for generating user preferences,
The determining means is characterized in that the content corresponding to the user preference is set as a content to be reproduced next in a certain cycle.

The information processing apparatus according to claim 4 is:
The reference is a certain range of similarity (eg, FIG. 3).
It is characterized by that.

The information processing apparatus according to claim 5 is:
The criterion is the probability that the content having a similarity higher than a predetermined level with the content whose playback order has been determined will be played next (for example, FIG. 6).
It is characterized by that.

The information processing apparatus according to claim 6 is:
The determination means sets the content selected by the user as the content to be reproduced first (for example, the reproduction music determination unit 25 in FIG. 2 that performs the processing in steps S2 and S3 in FIG. 4).
It is characterized by that.

An information processing apparatus according to claim 7
It further comprises user preference generation means (for example, user preference vector generation unit 22 in FIG. 2) for generating user preferences,
The determining means determines the content to be reproduced first based on the similarity between the user preference and the reproducible content (for example, the reproduction in FIG. 2 which performs the processing in steps S2, S5 and S6 in FIG. 4). Music determination unit 25)
It is characterized by that.

An information processing apparatus according to claim 8 is provided.
It further comprises user preference generation means (for example, user preference vector generation unit 22 in FIG. 2) for generating user preferences,
The content determining means is characterized in that the content to be reproduced next is determined from the contents corresponding to the reference with the similarity calculated by the calculating means and corresponding to the user preference.

An information processing apparatus according to claim 9 is provided.
Further comprising separation plane generation means (for example, a separation plane generation unit 31 in FIG. 5) for generating a separation plane that likes and dislikes from user preferences,
The content determining means is characterized in that the content to be reproduced next is determined from the contents corresponding to the reference and having the similarity calculated by the calculating means and belonging to the favorite plane of the separation plane.

An information processing apparatus according to claim 10 is provided.
The content determination means includes a content having a musical feature that is most musically continuous with the content whose playback order is determined, among the contents whose similarity calculated by the calculation means corresponds to the reference. The content to be played back (for example, the music having the first few bars whose music features of the last few bars of the previous music are most similar)
It is characterized by that.

The information processing method according to claim 11 and the program according to claim 12 are:
A calculation step (for example, step S10 in FIG. 4) for calculating the degree of similarity between the content whose playback order is determined and other content that can be played back;
Based on the similarity calculated in the processing of the calculation step, the content is reproduced next to the content for which the reproduction order is determined so that the similarity between the two content that is continuously reproduced changes with a certain fluctuation. And a determination step for determining content (for example, step S13 and step S14 in FIG. 4).

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 shows a configuration example of a playback apparatus 1 to which the present invention is applied.

  The operation input unit 11 of the playback device 1 is configured by an input device such as a touch panel or a cross key, for example, and notifies the control unit 12 of the operation performed by the user.

  The control unit 12 controls each unit. For example, the control unit 12 controls the playlist generation unit 16 to generate a playlist, or controls the playback unit 13 to continuously play back music in that order according to the generated playlist.

  Under the control of the control unit 12, the reproduction unit 13 appropriately reads and reproduces music data from the music data storage unit 14, and outputs the audio data obtained as a result from the output unit 15 including a speaker or the like.

  The music data storage unit 14 stores music data of reproducible music.

  The playlist generation unit 16 determines a song whose similarity with the music played back one time (the music whose playback order has been determined) becomes a similarity corresponding to a predetermined criterion as a music to be played next. A playlist is generated, but by changing the criteria as appropriate, the similarity between two pieces of content that are continuously played is changed with a certain fluctuation.

  FIG. 2 shows a configuration example of the playlist generation unit 16.

  The music vector storage unit 21 stores music vectors generated from the metadata of music (reproducible music) stored in the music data storage unit 14 (FIG. 1).

The music vector is, for example, a k-dimensional vector as shown below in which music attribute items such as “artist name”, “tempo”, “genre”, and “rhythm” are elements VA.
Music vector = {VA1, VA2,..., VAk}

The element VAi (i = 1, 2,..., K) is a vector composed of a plurality of elements c corresponding to the attribute items as shown below, and the value of the element c depends on the attribute. It is determined.
VAi = (c1, c2,...)

For example, the “tempo” vector VA as the element VA is composed of elements c of “fast”, “slightly fast”, “normal”, “slightly slow”, and “slow” as shown below. The “tempo” vector VA of a music with a fast speed is as follows (the value of the element c corresponding to the attribute is 1 and the values of the other elements c are 0).
“Tempo” vector VA = (fast, slightly fast, normal, slightly slow, slow)
"Tempo" vector VA = (1, 0, 0, 0, 0)

The “genre” vector VA as the element VA is composed of the elements c of “pops”, “jazz”, “classic”, “techno”,... The “genre” vector VA is as shown below.
"Genre" vector VA = (pops, jazz, classic, techno, ...)
"Genre" vector VA = (1, 0, 0, 0, ...)

Eventually, for example, the music vector of a music having a fast “tempo” and a pop of “genre” is as follows.
Music vector = {.., "tempo" VA (1,0,0,0,0), ..., "genre" VA (1,0,0,0, ...), ...

  The user preference vector generation unit 22 acquires the music vector of the reproduced music from the music vector storage unit 21 and updates the user preference vector stored in the user preference vector storage unit 23 based on the music vector. Thus, user preference information is accumulated. Specifically, each vector VA constituting the music vector of the music just reproduced is added to each vector VA constituting the user preference vector currently stored in the user preference vector storage unit 23.

For example, the music vector of the played music is
{"Tempo" VA (0,0,0,0,1), ..., "Genre" VA (1,0,0,0, ...),
The user preference vector stored in the user preference vector storage unit 23 is
{･ ･ "Tempo" VA (16,8,0,8,4), ..., "Genre" VA (10,8,5,1, ...), ...}, the user preference vector is ,
{·· “tempo” VA (16,8,0,8,5),..., “Genre” VA (11,8,5,1,...),.

  In this example, the music vector is stored in advance in the music vector storage unit 21, and the user preference vector generation unit 22 acquires the music vector from the music vector storage unit 21. If it is stored in the data storage unit 14 in correspondence with the music data, its metadata) can be acquired as appropriate, and a music vector can be generated accordingly.

The similarity calculation unit 24 calculates the similarity between a predetermined music piece and other music pieces based on those music vectors, as shown in Expression (1), in response to a request from the playback music determination unit 25. Or a similarity is calculated from a predetermined music vector and a user preference vector, and is supplied to the reproduction music determination unit 25 as appropriate.

  The reproduction music determination unit 25 has a similarity (reference similarity) corresponding to a predetermined reference in the reference table (FIG. 3) in which the similarity to the music reproduced immediately before is stored in the reference table storage unit 26. Is determined as the music to be reproduced next, and the reference is changed every time the music to be reproduced is determined.

  FIG. 3 shows an example of a reference table stored in the reference table storage unit 26. This reference table is composed of reference numbers 1 to m (= 8) numbered in order from the top in the figure and a predetermined similarity range associated therewith.

That is, when the playback music is determined while changing the reference based on the reference table of FIG.
The first song (the first song in the playlist) is determined, for example, by the user's song selection,
The second music reproduced next to the first music is a music whose similarity with the first music is a similarity within the range of 0.8 to 1.0 (standard similarity range of reference number 1). And
The third music is a music whose similarity with the second music is 0.8 to 1.0 (standard similarity range of reference number 2),
The fourth music is a music whose similarity with the third music is 0.6 to 0.8 (standard similarity range of reference number 3),
The fifth music is a music whose similarity to the fourth music is 0.4 to 0.6 (standard similarity range of reference number 4),
The sixth music is a music whose similarity with the fifth music is 0 to 0.2 (standard similarity range of reference number 5),
The seventh music is a music whose similarity with the sixth music is 0.8 to 1.0 (standard similarity range of reference number 6),
The eighth music is a music whose similarity to the seventh music is 0.6 to 0.8 (standard similarity range of reference number 7),
The ninth song is a song whose similarity with the eighth song is 0.8 to 1.0 (standard similarity range of reference number 8),
The determination of the tenth song is again a song whose similarity with the ninth song is a similarity in the range of 0.8 to 1.0 based on the reference similarity range of reference number 1. .
The eleventh and subsequent tunes are also tunes whose similarity to the previously played tune is similar to the criteria changed in this way.

  That is, the playback order is determined so that the similarity between two content played back continuously changes with a certain fluctuation.

  Next, the playlist generation process will be described with reference to the flowchart of FIG.

  In step S1, when a power source (not shown) of the playback device 1 is turned on, in step S2, the playback music determination unit 25 of the playlist generation unit 16 operates the operation input unit 11 to be played back by the user. Is determined for a predetermined time, for example, based on the control signal from the control unit 12, and if it is determined that the music has been selected during that time, the process proceeds to step S3.

  In step S <b> 3, the reproduction music determination unit 25 sets the selected music as the music to be reproduced first (the first music in the playlist).

  In step S4, the reproduction music determination unit 25 determines a similarity criterion for determining the music to be reproduced second. In this case, the playback music determination unit 25 acquires the reference number 1 using the reference (FIG. 3) of the reference number 1 in the reference table as the reference for determining the second music.

  On the other hand, if it is determined in step S2 that there is no music selection, the process proceeds to step S5, and the reproduction music determination unit 25 notifies the similarity calculation unit 24 to that effect. The similarity calculation unit 24 reads the user preference vector from the user preference vector storage unit 23, reads each music vector from the music vector storage unit 21, and calculates the similarity from the user preference vector and each music vector (formula (1)), supplied to the playback music determination unit 25.

  Next, in step S6, the reproduction music determination unit 25 determines the music to be reproduced first (the first music in the playlist). In this case, the reproduction music determination unit 25 first reproduces the music (the music that is most likely to match the user's preference) that has the highest similarity among the similarities supplied from the similarity calculation unit 24. The music to be played.

  In step S <b> 7, the reproduction music determination unit 25 determines a similarity criterion for determining the music to be reproduced second. In this case, the reproduction music determination unit 25 uses the reference of the reference number 1 in the reference table as a reference for determining the second music, and acquires the reference number 1. Note that the playback music determination unit 25 can detect a reference corresponding to the highest similarity detected in step S6 from the reference table, and use the reference as a reference when determining the second music. .

  In step S4 or step S7, when the reference for determining the second music is determined, the playback music determination unit 25 sets the value of the counter i indicating 1 to which the playback order has been determined to 1 Set (in this case, the value of counter i is 1 because only the first music to be played is determined).

  Next, in step S9, the reproduction music determination unit 25 sets the reference number determined in step S4 or step S7 in the counter j. In this case, the reference number 1 is set to the counter j.

  Next, it progresses to step S10 and the music reproduced | regenerated after the 2nd is determined by the process after it.

  In step S10, the degree of similarity between the music played in the order indicated by the counter i (hereinafter referred to as the i-th music) (in this case, the first music) and the other music is calculated. . Specifically, the reproduction music determination unit 25 notifies the similarity calculation unit 24 of the first music. The similarity calculation unit 24 reads the music vector of the first music and the other music notified from the reproduction music determination unit 25 from the music vector storage unit 21, and uses them as the first music and The similarity with other music is calculated. The similarity calculation unit 24 supplies the calculated similarity to the reproduction music determination unit 25.

  In step S11, the reproduction music determination unit 25 uses the reference number indicated by the counter j to indicate the similarity to the i-th music from the similarity to the i-th music supplied from the similarity calculation unit 24 in step S10. It is determined whether or not there is a music piece having a similarity within the similarity range indicated by the reference (hereinafter referred to as a reference of the reference number j). If it is determined that there is a song, the process proceeds to step S12.

  In this case, it is determined whether or not there is a song whose similarity with the first song is a similarity between 0.8 and 1.0 indicated by the reference of reference number 1.

  In step S12, the reproduction music determination unit 25 determines that the music whose similarity to the i-th (in this case, the first) music is a similarity within the similarity range indicated by the reference number j is 2 If it is determined whether there are two or more, and it is determined that there are two or more, the process proceeds to step S13, the music having the highest similarity is selected, and the music to be reproduced next (the first music) 2nd music).

  When it is determined in step S12 that the number of similarities with the i-th (in this case, the first) music is similar within the similarity range indicated by the criterion of the reference number j is one. In step S14, the reproduction music determination unit 25 sets the music as a music to be reproduced next (second music).

  When the next music to be reproduced (in this case, the second music) is determined in step S13 or step S14, the process proceeds to step S15, and the reproduced music determination unit 25 determines whether the counter i> = n ( If it is determined whether the reproduction order of all n pieces of music for which the reproduction order is determined has been determined) and it is determined that the counter i> = n is not satisfied (there are undecided music remaining), the process proceeds to step S16. move on.

  In step S <b> 16, the playback music determination unit 25 increments the value of the counter i by 1.

  When the value of the counter i is incremented by 1 in step S16, or in step S11, there is no song whose similarity is within the similarity range indicated by the criterion of the reference number j with the i-th song. If it is determined, the process proceeds to step S17.

  In step S17, the reproduced music determination unit 25 determines whether or not the counter j> = m (m = 8 in the example of FIG. 3) (whether the reference of the last reference number has been referenced). When it is determined that the counter j> = m is not satisfied, the process proceeds to step S18, and the value of the counter j is incremented by 1 (the reference for determining the next music to be played is changed).

  When it is determined in step S17 that the counter j> = m, the process proceeds to step S19, and the reproduction music determination unit 25 sets the counter j = 1 (the reference for determining the music to be reproduced next is the reference It is changed to the reference number 1 (returned)).

  When the value of the counter j is changed in step S18 or step S19, the process returns to step S10, and the subsequent processing is executed.

  That is, when it is determined that there is a song whose similarity with the first song is within the similarity range indicated by the reference of reference number 1 (step S11), the first song is set at step S13 or step S14. The second music is determined from the music whose similarity to the second music is the similarity within the similarity range indicated by the reference of reference number 1.

  On the other hand, when there is no music whose similarity with the first music is within the similarity range indicated by the reference of the reference number 1 (step S11), the similarity reference is the reference of the reference number 2. It is changed (step S18), and the second music is determined from the music whose similarity with the first music is the similarity within the similarity range indicated by the reference number 2. That is, the similarity criterion is changed until a song whose similarity with the first song is within the similarity range indicated by the criterion is obtained.

  Similarly, the music after the third music is determined by repeatedly executing the processes of Step S10 to Step S19.

  When it is determined in step S15 that the counter i> = n (when the playback order of all n pieces of music has been determined), the process proceeds to step S20, and the playlist generating unit 16 determines as described above. A playlist is generated based on the playback order.

  As described above, a playlist is generated in which the degree of similarity between two pieces of music that are continuously played changes with a certain fluctuation. That is, according to this playlist, the similarity between songs that are continuously played changes in this way, and songs with the same musical tone are not played continuously, so that it is possible to continuously play songs without getting tired. .

  By the way, the music to be reproduced next is determined based on the degree of similarity with the music reproduced immediately before, so that it may be considered that the music that is different from the user's preference is selected. Therefore, the music set in the playlist can be set as a music within a certain range corresponding to the user preference to some extent.

  A configuration example of the playlist generation unit 16 in this case is shown in FIG. The playlist generation unit 16 is further provided with a separation plane generation unit 31 in addition to the playlist generation unit 16 of FIG.

  The separation plane generating unit 31 generates a favorite and disliked separation plane by SVM (Support Vector Machine) or the like from the user preference vector stored in the user preference vector storage unit 23, and the song vector stored in the song vector storage unit 21. Among them, the music vector of the music classified as favorite is extracted and supplied to the similarity calculation unit 24.

  That is, in step S10 in FIG. 4, the similarity calculation unit 24 calculates the similarity between the i-th song and other songs, and the music vector (classified as favorite) supplied from the separation plane generation unit 31. Each similarity is calculated using only the music vector of the music that has been played. As a result, the music to be reproduced next is determined from the music classified as favorite (step S13 or step S14), so that the music that is distant from the user's preference is not set in the playlist. .

  In addition, as a method of preventing a song that is distant from the user preference from being set in the playlist, after determining the playback order of a predetermined number of songs, the music that best matches the user preference among the remaining songs is The music may be reproduced (for example, the flow returns to step S5 when the value of the counter i reaches a predetermined value). By doing in this way, since it returns to the music suitable for a user's preference for every some music, it can prevent that the music away from a user preference is set.

In the present example, a reference table is prepared and the similarity criterion is changed, but instead of the reference table, the reference similarity range is calculated every time using equation (2). It can also be done.

  Further, in place of the reference table of FIG. 3, as shown in FIG. 6, the similarity with the music reproduced one before becomes a certain degree of similarity (0.8 or more in the example of FIG. 6). It is also possible to set the probability that the music will be played next. If the playback order is determined using the reference table of FIG. 6, there is a high possibility that a music similar to the previous music will be played at first, but it will gradually decrease and increase again. Music can be reproduced.

  In the case of this example, for example, as the processing of step S11 in FIG. 4, the reproduction music determination unit 25 has a probability that the music whose similarity to the i-th music is 0.8 or more is indicated by the reference of the reference number j The music to be reproduced next is determined so as to be reproduced next in FIG.

  In the above description, when there are a plurality of music pieces belonging to the reference, the music piece having the highest similarity is selected (step S13 in FIG. 4), but the music feature amount (rhythm, tempo, etc.) of the last few bars of the previous music piece is selected. It is also possible to select a song with the first few bars that are most similar. By doing so, the connection of songs can be made musically natural.

  FIG. 7 shows another configuration example of the playlist generation unit 16. The playlist generation unit 16 is provided with a reproduction music determination unit 51 instead of the reproduction music determination unit 25 of FIG. The other parts are the same as those in FIG.

  The playback music determination unit 51 in FIG. 7 determines a music whose similarity to the user preference is a similarity corresponding to a predetermined standard as a music to be played next, and generates a playlist. As a result, the similarity between the content to be reproduced and the user preference is changed with a certain fluctuation.

That is, when the playback music is determined while changing the reference based on the reference table of FIG.
The first song (the first song in the playlist) is determined, for example, by the user's song selection,
The second music that is played after the first music is a music that has a similarity with the user preference vector of 0.8 to 1.0 (reference similarity range of reference number 1). And
The third song is a song whose similarity with the user preference vector is a similarity in the range of 0.8 to 1.0 (reference similarity range of reference number 2),
The fourth song is a song whose similarity with the user preference vector is a similarity in the range of 0.6 to 0.8 (reference similarity range of reference number 3),
The fifth song is a song whose similarity with the user preference vector is a similarity in the range of 0.4 to 0.6 (reference similarity range of reference number 4),
The sixth music is a music whose similarity with the user preference vector is a similarity in the range of 0 to 0.2 (reference similarity range of reference number 5),
The seventh song is a song whose similarity with the user preference vector is a similarity in the range of 0.8 to 1.0 (reference similarity range of reference number 6),
The eighth music is a music whose similarity with the user preference vector is 0.6 to 0.8 (standard similarity range of reference number 7),
The ninth song is a song whose similarity with the user preference vector is a similarity in the range of 0.8 to 1.0 (reference similarity range of reference number 8),
The determination of the tenth music piece is again a music piece whose similarity with the user preference vector is within the range of 0.8 to 1.0 based on the reference similarity range of reference number 1.
The eleventh and subsequent music pieces are also music pieces whose similarity with the user preference vector is the degree of similarity according to the criteria changed in this way.

  This playlist generation process will be described with reference to the flowchart of FIG.

  When the playback apparatus 1 is turned on in step S31, the similarity calculation unit 24 reads the user preference vector from the user preference vector storage unit 23 and the music vector from the music vector storage unit 21 in step S32. The similarity is calculated from the user preference vector and each music vector and supplied to the reproduction music determination unit 51.

  In step S <b> 33, the reproduction music determination unit 51 determines, for example, for a predetermined time based on the control signal from the control unit 12 whether or not the music to be reproduced is selected by the operation of the operation input unit 11 by the user. If it is determined that there is a song selection during that time, the process proceeds to step S34.

  In step S34, the reproduced music determination unit 51 sets the selected music as the music to be reproduced first (the first music in the playlist).

  In step S <b> 35, the playback music determination unit 51 determines a similarity criterion when determining the second music to be played back. In this case, the reproduction music determination unit 51 notifies the similarity calculation unit 24 of the music to be reproduced first (the music selected by the user). The similarity calculation unit 24 calculates the similarity from the music vector of the first music and the user preference vector, and supplies the similarity to the reproduction music determination unit 51. The reproduction music determination unit 51 detects a reference to which the similarity supplied from the similarity calculation unit 24 of the reference table belongs, and uses the reference as a reference for determining the second music, and the reference number is used as the reference. get.

  For example, when the similarity calculated from the music vector of the first music and the user preference vector is 0.4, the reference of the reference number 4 (FIG. 3) is the reference for determining the second music. The reference number 4 is acquired.

  If it is determined in step S33 that there is no music selection, the process proceeds to step S36, where the reproduction music determination unit 51 calculates the similarity calculated from the user preference vector supplied from the similarity calculation unit 24 and each music vector in step S32. The music with the highest similarity is obtained as the music that is reproduced first (first music).

  In step S <b> 37, the playback music determination unit 51 determines the similarity criterion for determining the music to be played back second. In this case, the reproduction music determination unit 51 detects a reference to which the similarity (the highest similarity) detected in step S36 of the reference table belongs, and uses that as a reference for determining the second music. And obtain the reference number.

  When the reference for determining the second music is determined in step S35 or step S37, the process proceeds to step S38, and the reproduction music determination unit 51 sets the value of the counter i to 1.

  Next, in step S39, the reproduced music determination unit 51 sets the reference number determined in step S35 or step S37 in the counter j.

  Next, it progresses to step S40 and the music reproduced | regenerated after the 2nd is determined by the process after it.

  In step S40, the playback music determination unit 51 determines the similarity between the user preference vector and the similarity with the reference number j based on the similarity with the user preference vector supplied from the similarity calculation unit 24 in step S32. It is determined whether or not there is a song having a similarity within the range. If it is determined that there is a song, the process proceeds to step S41.

  In step S41, the i-th song (in this case, the first song) among the songs whose similarity to the user preference vector is the similarity within the similarity range indicated by the reference of the reference number j. It is determined whether or not there is a music piece that has a similarity degree equal to or greater than a certain degree.

  Specifically, in the reproduction music determination unit 51, the similarity between the i-th music (first music) and the user preference vector is a similarity within the similarity range indicated by the reference number j. The music (step S40) is notified to the similarity calculation unit 24, and the similarity calculation unit 24 calculates the similarity between the i-th music and those music and supplies it to the reproduction music determination unit 51. . The reproduction music determination unit 51 detects music whose similarity to the i-th music is equal to or higher than a predetermined similarity from the similarity supplied from the similarity calculation unit 24.

  In step S42, the playback music determination unit 51 determines the music that has been detected in the process of step S41 (similarity with the user preference vector is the similarity within the similarity range indicated by the criterion of the reference number j, and the i th It is determined whether or not there are two or more music pieces (songs in which the degree of similarity is equal to or higher than a certain degree) with the first music piece (in this case, the first music piece). The music with the highest similarity to the user preference vector is selected, and the music is set as the music to be reproduced next (in this case, the second music).

  In step S42, when one piece of music is detected in the process of step S41, the process proceeds to step S44, and the reproduction music determination unit 51 selects the music to be reproduced next (the second music in this case). ).

  When the next music to be reproduced (second music) is determined in step S43 or step S44, the process proceeds to step S45, and whether the counter i> = n is satisfied (the playback order of all n music is determined) If it is determined that the counter i> = n is not yet satisfied (there are undetermined songs remaining), the process proceeds to step S46.

  In step S46, the reproduced music determination unit 51 increments the value of the counter i by 1.

  When the value of the counter i is incremented by 1 in step S46, it is determined in step S40 that there is no song whose similarity with the user preference vector is within the similarity range indicated by the reference number j. Or in step S41, the similarity with the i-th music is more than a certain degree among the music whose similarity with the user preference vector is within the similarity range indicated by the reference of the reference number j When it is determined that there is no music to be used, the process proceeds to step S47.

  In steps S47 to S50, the same processing as in steps S17 to S20 of FIG. 4 is performed, and thus the description thereof is omitted.

  In this way, the second and subsequent music pieces are determined.

  In the present example, a reference table is prepared and the similarity criterion is changed, but instead of the reference table, the reference similarity range is calculated every time using equation (2). It can also be done.

  Further, in place of the reference table of FIG. 3, as shown in FIG. 6, the similarity with the music reproduced one before becomes a certain degree of similarity (0.8 or more in the example of FIG. 6). It is also possible to set the probability that the music will be played next. If the playback order is determined using the reference table of FIG. 6, there is a high possibility that a music similar to the previous music will be played at first, but it will gradually decrease and increase again. Music can be reproduced.

  In the case of the present example, for example, as the process of step S40 in FIG. 8, the reproduction music determination unit 25 indicates that the music whose similarity to the i-th music is 0.8 or more is indicated by the reference number j. The music to be played next is determined so as to be played next with probability (FIG. 6).

    In the above description, when there are a plurality of songs belonging to the reference, the song having the highest similarity is selected (step S43 in FIG. 8), but the music feature amount (rhythm, tempo, etc.) of the last few bars of the previous song is selected. It is also possible to select a song with the first few bars that are most similar. By doing so, the connection of songs can be made musically natural.

  In the example of FIG. 1, the case where a playlist is generated from a song in which the song data stored in the song data storage unit 14 is stored has been described as an example. However, as shown in FIG. Can be set as a playlist.

  In this case, the communication unit 61 of the playback device 1 communicates with the server 71 as appropriate, acquires music that can be provided to the user, and supplies it to the control unit 12. The control unit 12 stores the music supplied from the server 71 in the music data storage unit 14 and supplies the music vector of the music to the playlist generation unit 16. The playlist generation unit 16 generates a playlist as described above using the music vector supplied from the control unit 12.

  Moreover, although the example at the time of making a content into a music was demonstrated above, content other than a music may be sufficient as a content.

  The series of processes described above can be executed by dedicated hardware or can be executed by software. When a series of processing is performed by software, for example, the series of processing can be realized by causing a (personal) computer as shown in FIG. 10 to execute a program.

  In FIG. 10, a CPU (Central Processing Unit) 111 executes various processes according to a program stored in a ROM (Read Only Memory) 112 or a program loaded from a hard disk 114 to a RAM (Random Access Memory) 113. . The RAM 113 also appropriately stores data necessary for the CPU 111 to execute various processes.

  The CPU 111, the ROM 112, and the RAM 113 are connected to each other via a bus 115. An input / output interface 116 is also connected to the bus 115.

  The input / output interface 116 includes an input unit 118 including a keyboard, a mouse, and an input terminal, a display including a CRT (Cathode Ray Tube) and an LCD (Liquid Crystal display), an output terminal, and an output unit 117 including a speaker. A communication unit 119 including a terminal adapter, an ADSL (Asymmetric Digital Subscriber Line) modem, a LAN (Local Area Network) card, and the like is connected. The communication unit 119 performs communication processing via various networks such as the Internet.

  A drive 120 is also connected to the input / output interface 116, a magnetic disk (including a floppy disk) 131, an optical disk (including a CD-ROM (Compact Disk-Read Only Memory) DVD Digital Versatile Disk)) 132, a magneto-optical disk. 133 (including MD (Mini-Disk)) or a removable medium (recording medium) 134 such as a semiconductor memory is appropriately mounted, and a computer program read from the medium is installed in the hard disk 114 as necessary.

  In the present specification, the steps described in the flowcharts are executed in parallel or individually even if they are not necessarily processed in time series, as well as processes performed in time series in the described order. It also includes processing.

  Further, in this specification, the system represents the entire apparatus constituted by a plurality of apparatuses.

It is a block diagram which shows the structural example of the reproducing | regenerating apparatus 1 to which this invention is applied. It is a block diagram which shows the structural example of the play list production | generation part 16 of FIG. It is a figure which shows the example of the reference | standard table memorize | stored in the reference | standard table memory | storage part 26 of FIG. It is a flowchart explaining a play list generation process. It is a block diagram which shows the other structural example of the play list production | generation part 16 of FIG. It is a figure which shows the other example of the reference | standard table memorize | stored in the reference | standard table memory | storage part 26 of FIG. It is a block diagram which shows the other structural example of the play list production | generation part 16 of FIG. It is a flowchart explaining another playlist production | generation process. It is a block diagram which shows the other structural example of the reproducing | regenerating apparatus 1 to which this invention is applied. And FIG. 16 is a block diagram illustrating a configuration example of a personal computer.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Playback apparatus, 16 playlist production | generation parts, 21 music vector memory | storage part, 22 user preference vector production | generation part, 23 user preference vector memory | storage part, 24 similarity calculation part, 25 reproduction | regeneration music determination part, 26 reference | standard table memory | storage part, 31 separation Plane generation unit, 51 playback music determination unit, 61 communication unit, 71 server

Claims (12)

In an information processing apparatus for determining the playback order of one or more contents,
A calculating means for calculating the similarity between the content whose playback order has been determined and other content that can be played back;
Determining means for determining the playback order so that the similarity between two pieces of content that are played back continuously changes with a certain fluctuation based on the similarity calculated by the calculating means. A characteristic information processing apparatus. The determining means includes
Content determining means for determining a content to be played next to content whose playback order has been determined, from content whose similarity calculated by the calculating means is a similarity corresponding to a predetermined reference;
The information processing apparatus according to claim 1, further comprising: a changing unit that changes the reference so that the similarity between two pieces of content that are played back continuously changes with a certain fluctuation. It further comprises user preference generation means for generating user preferences,
The information processing apparatus according to claim 1, wherein the determination unit sets the content corresponding to the user preference as a content to be reproduced next at a constant cycle. The information processing apparatus according to claim 2, wherein the reference is a certain range of similarity. The information processing apparatus according to claim 2, wherein the criterion is a probability that a content having a similarity higher than or equal to a predetermined similarity with a content whose playback order has been determined is played back next. The information processing apparatus according to claim 1, wherein the determination unit sets the content selected by the user as the content to be reproduced first. It further comprises user preference generation means for generating user preferences,
The information processing apparatus according to claim 1, wherein the determination unit determines a content to be reproduced first based on a similarity between the user preference and a reproducible content. It further comprises user preference generation means for generating user preferences,
The content determination unit determines a content to be reproduced next from the content in which the similarity calculated by the calculation unit corresponds to the reference and corresponds to the user preference. 2. The information processing apparatus according to 2. Further comprising separation plane generating means for generating a separation plane that likes and dislikes from the user preference,
The content determination means determines the content to be reproduced next from the content whose similarity calculated by the calculation means corresponds to the reference and belongs to a favorite plane of the separation plane. The information processing apparatus according to claim 8. The content determination unit is a content having a musical feature that is most musically continuous with the content whose reproduction order is determined, among the content in which the similarity calculated by the calculation unit is a similarity corresponding to the reference. The information processing apparatus according to claim 2, wherein the information is the content to be reproduced next. In an information processing method for determining the playback order of one or more contents,
A calculation step for calculating the similarity between the content whose playback order has been determined and other content that can be played back;
Based on the similarity calculated in the process of the calculating step, the playback is performed after the content whose playback order is determined so that the similarity between the two content played back continuously changes with a certain fluctuation. An information processing method comprising: a determining step for determining content to be processed. In a program executed by a processor that controls an information processing apparatus that determines the playback order of one or more contents,
A calculation step for calculating the similarity between the content whose playback order has been determined and other content that can be played back;
Based on the similarity calculated in the process of the calculating step, the playback is performed after the content whose playback order is determined so that the similarity between the two content played back continuously changes with a certain fluctuation. And a determining step for determining the content to be played.

Images