JP6561786B2 - Broadcast program scene identification method, evaluation method, broadcast program scene identification device, and program - Google Patents

Description

  The present invention relates to a broadcast program scene specifying method, an evaluation method, a broadcast program scene specifying device, and a program.

  In the broadcasting station, after the broadcast program is broadcast, a digest version program (including program advertisement for the next broadcast) is created for the current broadcast. The digest version program is an opportunity for a series program viewer to watch the next broadcast. Therefore, the creation of a digest version program that suits the viewer's preference is also important from the viewpoint of enclosing the viewer.

JP 2009-200908 A

  However, conventionally, the creation of a digest version program is mainly performed based on the preference of the program organizer. Therefore, the content of the digest version program does not necessarily match the viewer's preference.

  In order to create a digest version program that matches the viewer's preference, it is considered important to identify a scene that is highly interested in the viewer in the broadcast program. That is, it is considered that the possibility of creating a digest version program that suits the viewer's preference is increased by including a scene that is highly interested in the viewer in the digest version program.

  For example, it is conceivable to identify a scene in which viewers are highly interested with reference to a time series of audience ratings measured for broadcast programs.

  However, the audience rating is an index measured in some homes, and is considered insufficient as information for grasping the general interest of the viewer. In particular, in recent years, there are many viewers who record and view broadcast programs and viewers via the Internet, and it is difficult to grasp scenes that are of great interest for these viewers in terms of audience rating.

  Therefore, an object of one aspect is to specify a scene of a broadcast program that is highly interested by viewers.

  In one plan, a scene identification method for a broadcast program refers to a first storage unit in which a posted message of a first information processing service is stored, and the first program is transmitted within a first period from the broadcast timing of the broadcast program. A second storage unit that identifies a keyword related to the broadcast program posted to the information processing service and stores a posted message of the second information processing service that is less immediate than the first information processing service. Referencing and specifying a keyword related to the broadcast program posted to the second information processing service within a second period longer than the first period from the broadcast timing of the broadcast program, and specifying each scene of the broadcast program The keyword posted to the specified first information processing service and the specified second information processing are referred to the third storage unit stored in association with the character string of Based on the post in the service keyword, among the broadcast programs, to identify the scene associated with at least one of the keywords identified, processing computer executes.

  According to one aspect, it is possible to specify a scene of a broadcast program that is highly interested by the viewer.

It is a figure which shows the system configuration example in embodiment of this invention. It is a figure which shows the hardware structural example of the scene specific device in embodiment of this invention. It is a figure which shows the function structural example of the scene specific device in embodiment of this invention. It is a flowchart for demonstrating an example of the process sequence of the acquisition process of audience rating data and a broadcast log. It is a figure which shows the structural example of audience rating data. It is a figure which shows the structural example of a broadcast log. It is a flowchart for demonstrating an example of the process sequence of the acquisition process of a posting message. It is a flowchart for demonstrating an example of the process sequence of the extraction process of the keyword from a posting message. It is a figure which shows an example of the extracted keyword. It is a flowchart for demonstrating an example of the process sequence of the allocation process of the appearance frequency with respect to each time. It is a figure which shows an example of the frequency | count of appearance for every unit time of each keyword according to SNS system. It is a figure which shows an example of the allocation result of the appearance coefficient with respect to each unit time. It is a figure which shows an example of the total result of the appearance frequency after correction | amendment. It is a figure which shows an example of broadcast progress data. It is a figure which shows the structural example of video material information. It is a figure which shows an example of the frequency | count of appearance allocated at each time based on the matching of a keyword and metadata. It is a figure which shows an example of the frequency | count of appearance allocated at each time based on the matching of a keyword and a broadcast log. It is a flowchart for demonstrating an example of the process sequence of the specific process of a scene. It is a figure which shows an example of the calculation result of the correction value for every time. It is a figure which shows an example of the audience rating data after correction | amendment. It is a flowchart for demonstrating an example of the process sequence performed regarding distribution of a digest version program. It is a figure which shows the structural example of viewing history data. It is a figure which shows an example of the calculation result of the frequency | count of viewing.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing an example of a system configuration in the embodiment of the present invention. In FIG. 1, the scene specifying device 10 is connected to the viewing information providing device 20, the SNS system 30a, and the SNS system 30b via a network such as the Internet.

  The scene identification device 10 is one or more computers that identify a scene that is highly interested in viewers in a broadcast program broadcast on a television and supports creation of a digest version program that matches the viewer's preference. The scene specifying device 10 is used by, for example, a broadcasting station or a broadcast program production company. In the present embodiment, a broadcast program broadcast on a specific date and time (hereinafter referred to as “program M”) is a digest version program creation target.

  The viewing information providing device 20 is one or more computers that provide the viewing rate, broadcast log data, and the like of broadcast programs related to television broadcasting to the scene specifying device 10. For example, audience ratings, broadcast log data, and the like are measured or generated by Video Research. Broadcast log data is data in which a history of the content actually broadcast at a certain time is recorded as a character string. The audience rating, broadcast log data, and the like are used to specify a scene that is highly interested by the viewer.

  The SNS system 30a and the SNS system 30b are computer systems for providing SNS (Social Networking Service), respectively. In the present embodiment, the message length that can be posted (writable) to the SNS system 30b is longer than the message length that can be posted to the SNS system 30a. Therefore, the service provided by the SNS system 30b has a lower immediacy regarding posting posted messages (write information) than the service provided by the SNS system 30a. Immediateness related to posting of a posted message refers to the length of time that elapses from when a certain event occurs until the posted message is posted. That is, the elapsed time from the occurrence of a certain event to the posting of a posted message tends to be shorter in the SNS system 30a and longer in the SNS system 30b. For example, the service provided by the SNS system 30a may be Twitter (registered trademark) or the like, and the service provided by the SNS system 30b may be Facebook (registered trademark), a blog, or the like. In the present embodiment, a posted message posted to the SNS system 30a or SNS system 30b during or after the broadcast of the broadcast program is used to identify a scene that is highly interested by the viewer. In addition, when not distinguishing SNS system 30a and SNS system 30b, it is called "SNS system 30".

  FIG. 2 is a diagram illustrating a hardware configuration example of the scene specifying device according to the embodiment of the present invention. 2 includes a drive device 100, an auxiliary storage device 102, a memory device 103, a CPU 104, an interface device 105, and the like that are mutually connected by a bus B.

  A program that realizes processing in the scene specifying apparatus 10 is provided by the recording medium 101. When the recording medium 101 on which the program is recorded is set in the drive device 100, the program is installed from the recording medium 101 to the auxiliary storage device 102 via the drive device 100. However, the program need not be installed from the recording medium 101 and may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program and also stores necessary files and data.

  The memory device 103 reads the program from the auxiliary storage device 102 and stores it when there is an instruction to start the program. The CPU 104 executes a function related to the scene specifying device 10 in accordance with a program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network.

  An example of the recording medium 101 is a portable recording medium such as a CD-ROM, a DVD disk, or a USB memory. An example of the auxiliary storage device 102 is an HDD (Hard Disk Drive) or a flash memory. Both the recording medium 101 and the auxiliary storage device 102 correspond to computer-readable recording media.

  FIG. 3 is a diagram illustrating a functional configuration example of the scene identification device according to the embodiment of the present invention. In FIG. 3, the scene identification device 10 includes a viewing information acquisition unit 11, an SNS setting reception unit 12, a posted message acquisition unit 13, a keyword extraction unit 14, an appearance count allocation unit 15, a scene identification unit 16, a viewing history acquisition unit 17, It has an age estimation unit 18 and a digest distribution unit 19. Each of these units is realized by processing that one or more programs installed in the scene specifying apparatus 10 cause the CPU 104 to execute.

  The scene specifying device 10 also includes an audience rating data storage unit 121, a broadcast log storage unit 122, an SNS setting storage unit 123, an acquisition message storage unit 124a, an acquisition message storage unit 124b, a keyword storage unit 125, and a broadcast progress data storage unit 126. The video material information storage unit 127, the appearance count storage unit 128 for each time, the digest storage unit 129, the viewing history storage unit 130, and the like are used. Each of these storage units can be realized using, for example, the auxiliary storage device 102 or the memory device 103, or a storage device that can be connected to the scene specifying device 10 via a network.

  The viewing information acquisition unit 11 acquires from the viewing information providing apparatus 20 viewing rate data indicating the viewing rate related to the program M in time series and a broadcast log of the program M. The viewing information acquisition unit 11 stores the acquired audience rating data in the audience rating data storage unit 121, and stores the acquired broadcast log in the broadcast log storage unit 122.

  The SNS setting accepting unit 12 accepts a setting of information related to the SNS system 30 from which the posted message is acquired from the user. For example, the SNS setting accepting unit 12 includes identification information of each SNS system 30 from which a posted message is acquired and information indicating classification according to immediacy regarding posting of the posted message (hereinafter referred to as “immediate classification information”). And accept. The immediacy classification information is information indicating that the immediacy is relatively high or relatively low, for example. However, the immediacy may be classified into three or more. In the present embodiment, the identification information of each of the SNS system 30a and the SNS system 30b is set as the acquisition destination of the posted message. Moreover, about the immediacy classification information regarding the SNS system 30b, a value indicating that immediacy is lower than the immediacy indicated by the immediacy classification information regarding the SNS system 30a is set. The set information is stored in the SNS setting storage unit 123.

  The posted message acquisition unit 13 includes a posted message storage unit 31a of the SNS system 30a set as an acquisition destination of the posted message and a posted message storage unit 31b of the SNS system 30b within a predetermined period from the broadcast of the program M. Get posted message group. That is, a group of posted messages posted to the SNS system 30a or the SNS system 30b is stored in the posted message storage unit 31a and the posted message storage unit 31b.

  In acquiring the posted message group, the posted message posted within a longer period from the broadcast timing of the program M is acquired from the SNS system 30b having relatively low immediacy. The posted message acquisition unit 13 stores the posted message acquired from the SNS system 30a in the acquired message storage unit 124a, and stores the posted message acquired from the SNS system 30b in the acquired message storage unit 124b. Note that, when the acquired message storage unit 124a and the acquired message storage unit 124b are not distinguished, they are referred to as an acquired message storage unit 124.

  The keyword extraction unit 14 extracts one or more character strings (hereinafter referred to as “keywords”) having a high appearance frequency (appearance frequency) from the posted message group stored in the acquired message storage unit 124, the appearance frequency of each keyword, and the like. To extract. The extraction result is stored in the keyword storage unit 125.

  The appearance count assigning unit 15 assigns the number of appearances of keywords related to the time to each time when the audience rating is recorded in the audience rating data stored in the audience rating data storage unit 121. Whether or not each keyword is related to each time is determined based on broadcast progress data stored in the broadcast progress data storage unit 126 and video material information stored in the video material information storage unit 127. Broadcast progress data is also called cue sheet, and is created when a broadcast program is produced. Broadcast progress data indicates time table information of one or more video files constituting a broadcast program. Specifically, the broadcast progress data includes information indicating the start time and the end time of each of one or more video files constituting the broadcast program. Video material information is used for each video file of the broadcast progress data, and for each of one or more video material files that became the material of the video file, any of the scenes included in the video material file is used for the video file. Contains information indicating whether or not That is, when a broadcast program is produced, first, a plurality of video material files are created by shooting a video. Video files are produced by extracting and editing necessary scenes from each video material file. The video material information includes a character string that characterizes the scene used from each video material file. A scene related to the keyword can be specified by matching the character string and the keyword, and the time related to the keyword can be specified by specifying the broadcast time of the scene based on the broadcast progress data. can do.

  In addition, the number of appearances assigned to each time decreases as the elapsed time (delay period) from the broadcast timing of the program M increases so that the relationship between the program M and the keyword is reflected. Weighting is performed. At this time, when a certain keyword is detected from the posted message group acquired from the SNS system 30b, the posting is performed from the broadcast timing as opposed to the case where the keyword is detected from the posted message group acquired from the SNS system 30a. An evaluation algorithm is applied so that the decrease in weight according to the increase in the delay period until the timing becomes slow. This is because the posting of the posting message to the SNS system 30b has a relatively low immediacy, and is considered to be highly likely to be posted for a longer period from the broadcast of the program M.

  The weighted appearance count assigned for each time is stored in the appearance count storage section 128 for each time. Note that the appearance count assigning unit 15 separately assigns a weighted appearance count of a keyword related to the time to each time using the broadcast log stored in the broadcast log storage unit 122. In this case, the broadcast log and the keyword are matched, and the weighted number of appearances of the keyword is assigned to the time corresponding to the broadcast log including a certain keyword. Therefore, the number-of-appearance count storage unit 128 stores the number of appearances assigned to each time due to the match between the broadcast progress data and the keyword and the number of appearances assigned to each time due to the match between the broadcast log and the keyword. Is done.

  The scene specifying unit 16 includes a time-series audience rating of the program M indicated by the audience rating data stored in the audience rating data storage unit 121, and the number of appearances stored in the number-of-time appearance count storage unit 128 for each time. Based on the above, a scene that is estimated to be highly interested by the viewer is identified, and information indicating the scene is output. The producer of the digest version program produces a digest version program of program M with reference to the output information. The digest version program is produced according to the age of viewers. The produced digest version program is stored in the digest storage unit 129 by age.

  The viewing history acquisition unit 17 acquires viewing history data of an internet-connected television by a certain viewer from the viewing information providing device 20. The viewing history data is information indicating the channels viewed by the viewer over a period of several days in time series. The age estimating unit 18 compares the viewing history data with the audience rating data and estimates the age of the viewer. The digest distribution unit 19 distributes a digest version program corresponding to the age estimated by the age estimation unit 18 to the viewer.

  Hereinafter, a processing procedure executed by the scene specifying device 10 of FIG. 1 will be described. FIG. 4 is a flowchart for explaining an example of a processing procedure of acquisition processing of audience rating data and broadcast log.

  In step S <b> 101, the viewing information acquisition unit 11 acquires viewing rate data regarding the program M from the viewing information providing apparatus 20. Subsequently, the viewing information acquisition unit 11 stores the audience rating data in the audience rating data storage unit 121 (S102).

  FIG. 5 is a diagram illustrating a configuration example of audience rating data. As shown in FIG. 5, the audience rating data includes audience ratings by age group in time series. Specifically, the audience rating data shown in FIG. 5 includes audience ratings for each age group of teens, 20s, 30s, 40s, 50s, 60s, and 70s or more at 1-minute intervals. It is included. It is assumed that the broadcast date and time of program M is from 8:00 to 8:15 on January 23. In FIG. 5, the channel number “101” is the number of the channel on which the program M was broadcast. The channel number is a channel identification number on the broadcasting station side, and is not necessarily a television channel number on the viewer side. In FIG. 5, the ellipse surrounding a part of the audience rating is for later description.

  Subsequently, the viewing information acquisition unit 11 acquires a broadcast log related to the program M from the viewing information providing apparatus 20 (S103). Subsequently, the viewing information acquisition unit 11 stores the broadcast log in the broadcast log storage unit 122 (S104).

  FIG. 6 is a diagram illustrating a configuration example of a broadcast log. As shown in FIG. 6, the broadcast log includes the broadcast contents of the program M in time series. Each record of the broadcast log includes a broadcast date, a start time, an end time, corner details, and the like. The broadcast date is the broadcast date of program M. The start time and the end time are a start time and an end time for one corner among a plurality of corners constituting the program M. The break of the corner is determined by, for example, the creator of the broadcast log. For example, if the target program of the broadcast log is a drama, the corner may be divided for each scene of the drama. If the target of the broadcast log is a variety program, the corner of the broadcast log may be divided for each corner of the program. The corner details are character strings indicating the contents of one corner.

  Then, the process procedure which the SNS setting reception part 12 and the posting message acquisition part 13 perform is demonstrated. FIG. 7 is a flowchart for explaining an example of the processing procedure of the posted message acquisition processing.

  In step S <b> 111, the SNS setting reception unit 12 receives a setting of information related to the SNS system 30 from the user of the scene identification device 10. Specifically, the identification information (hereinafter referred to as “SNSID”) of the SNS system 30 from which the posted message is acquired and the immediacy classification information of each SNS system 30 are received. In the present embodiment, the SNS ID and immediacy classification information of the SNS system 30a and the SNS ID and immediacy classification information of the SNS system 30b are set.

  Subsequently, the SNS setting reception unit 12 stores the received setting content in the SNS setting storage unit 123 (S112). Therefore, the SNS setting storage unit 123 stores the SNS ID and the immediacy classification information for each SNS system 30.

  In step S <b> 113, the posted message acquisition unit 13 acquires a posted message group from the SNS system 30 corresponding to each SNS ID stored in the SNS setting storage unit 123. Note that step S113 may be executed asynchronously with steps S111 and S112. That is, steps S111 and S112 may be performed in advance.

  In the present embodiment, a posted message group is acquired from the posted message storage unit 31a of the SNS system 30a, and a posted message group is acquired from the posted message storage unit 31b of the SNS system 30b. Hereinafter, when distinguishing the posted message acquired from each SNS system 30, the posted message acquired from the SNS system 30a is referred to as “posted message A”, and the posted message acquired from the SNS system 30b is referred to as “posted message B”. That's it.

  The post message group to be acquired may be limited to those posted within a predetermined period from the broadcast timing (broadcast start time) of the program M, for example. The posted message acquisition unit 13 may change the length of the predetermined period based on the immediacy classification information stored in the SNS setting storage unit 123 for each SNS system 30. That is, the said predetermined period regarding the SNS system 30b with low immediacy may be made longer than the said predetermined period regarding the SNS system 30a with high immediacy. This is because the postal message related to the program M is more likely to be posted in the SNS system 30b having a lower immediacy in a longer time after the program M ends. For example, a post message posted within 1 hour and 15 minutes from the start of broadcasting of the program M is acquired from the SNS system 30a, and posted within 2 hours from the start of broadcasting of the program M from the SNS system 30b. A posted message may be acquired.

  Subsequently, the posted message acquisition unit 13 stores the acquired posted message group in the acquired message storage unit 124 (S114). In the present embodiment, the posted message A group is stored in the acquired message storage unit 124a, and the posted message B group is stored in the acquired message storage unit 124b.

  Next, an example of a processing procedure executed by the keyword extracting unit 14 for the posted message group acquired by the posted message acquiring unit 13 will be described. FIG. 8 is a flowchart for explaining an example of a processing procedure for extracting a keyword from a posted message.

  The keyword extraction unit 14 analyzes the posted message A group stored in the acquired message storage unit 124a and the posted message B group stored in the acquired message storage unit 124b (S121), and appears in these posted message groups. A keyword having a relatively large number of times and the number of appearances of the keyword are extracted (S122). A known document analysis technique may be used for analyzing the posted message group. For example, Japanese Patent No. 5245255, Japanese Patent No. 4347226, Japanese Patent No. 5206197, Japanese Patent No. 5206196, Japanese Patent No. 5245908, Japanese Patent No. 4804591, Japanese Patent No. 486078, or Japanese Patent No. 495779 The technique described in the above may be used.

  Subsequently, the keyword extraction unit 14 stores the extracted keywords in the keyword storage unit 125 (S123). For example, the keywords are sorted in descending order of the number of appearances and stored in the keyword storage unit 125.

  FIG. 9 is a diagram illustrating an example of the extracted keywords. In FIG. 9, the result sorted in descending order of the number of appearances is shown.

  By the way, there is a possibility that many posting messages unrelated to the program M may be posted on the SNS system 30. Therefore, in order to increase the possibility that keywords related to the program M are extracted, keywords may be extracted within the range of the broadcast log stored in the broadcast log storage unit 122. For example, the keyword extraction unit 14 may extract a word by analyzing a broadcast log, and may extract a keyword that appears frequently in the posted message group from the extracted word.

  Subsequently, a processing procedure executed by the appearance count assigning unit 15 will be described. FIG. 10 is a flowchart for explaining an example of the processing procedure of the appearance frequency assignment processing for each time.

  In step S <b> 201, the appearance count assignment unit 15 reads each keyword (FIG. 9) stored in the keyword storage unit 125. Subsequently, the appearance count allocation unit 15 acquires the appearance count per unit time for each keyword for each SNS system 30 (S202).

  FIG. 11 is a diagram illustrating an example of the number of appearances of each keyword per unit time for each SNS system. In FIG. 11, (1) indicates the number of appearances of each keyword per unit time in the posted message A group stored in the acquired message storage unit 124a. (2) indicates the number of appearances of each keyword per unit time in the posted message group B stored in the acquired message storage unit 124b.

  The number of appearances per unit time may be obtained by referring to the acquired message storage unit 124a and the acquired message storage unit 124b again in step S202, or calculated when a keyword is extracted in step S122 of FIG. May be. The number of appearances of a keyword per unit time refers to the number of appearances of the keyword in a posted message group posted within the unit time.

  In FIG. 11, 15 minutes is a unit time. Further, regarding (1), the number of appearances concerning five unit times is shown, and regarding (2), the number of appearances concerning eight unit times is shown. That is, the number of appearances is acquired over a long period of time for the posted message group B acquired from the SNS system 30b.

  Subsequently, the appearance count assigning unit 15 assigns an appearance coefficient to each unit time (S203). The appearance coefficient is a coefficient for weighting the number of appearances in each unit time. That is, the weighting is performed according to the possibility that the posted message posted in each time zone is related to the program M.

  FIG. 12 is a diagram illustrating an example of an appearance coefficient assignment result for each unit time. FIG. 12 shows the result of assigning the appearance coefficient to each unit time in FIG.

  Basically, since the viewer is watching the program while the program is being broadcast, it is considered that the viewer is more likely to post after the program ends than during the program. Therefore, in FIG. 12, the appearance coefficient is set larger after the end of the program than during the program (8: 0 to 8:15). Also, the longer the elapsed time from the end of the program, the fewer posted messages related to the program M. Therefore, in FIG. 12, the appearance coefficient is assigned so as to decrease as the elapsed time becomes longer. Also, in (2) rather than (1), each unit time is set so that the decrease in the appearance coefficient according to the increase in the delay period from the broadcast timing of the program M to the posting timing of the posted message becomes slower. An appearance coefficient is assigned to. This is because the SNS system 30 with relatively low immediacy is considered to have a longer delay period until posting of a posted message after the broadcast of the program M ends. In (1), the appearance coefficient after 9:15 is 0. In (2), the appearance coefficient after 10:00 is 0. Note that the keyword extraction period may be several days.

  Subsequently, the appearance count allocation unit 15 corrects the appearance count of each keyword in each unit time by the appearance coefficient for each SNS system 30 (S204). Specifically, for each keyword, a value obtained by multiplying the number of appearances of each unit time by the appearance coefficient is obtained. Subsequently, the appearance count allocation unit 15 counts the corrected (weighted) appearance count for each keyword (S205).

  FIG. 13 is a diagram illustrating an example of the result of counting the number of appearances after correction. FIG. 13 shows the total number of appearances after correction for each keyword in the rightmost columns of (1) and (2). In step S205, the total value for each SNS system 30 is added. As a result, a tabulation result as shown in (3) of FIG. 13 is obtained. In the following, the number of appearances for each keyword refers to the value shown in (3) of FIG.

  Subsequently, the appearance count allocating unit 15 reads the broadcast progress data of the program M stored in the broadcast progress data storage unit 126, and specifies the video file constituting the program M (S206).

  FIG. 14 is a diagram illustrating an example of broadcast progress data. In FIG. 14, the broadcast progress data includes the broadcast date, start time, end time, video file name, and the like for each video file used for broadcasting.

  The broadcast date is the broadcast date of the video file. The start time and end time are the broadcast start time and broadcast end time of the contents of the video file. The video file name is the file name of the video file.

  In step S205, a video file name including a start time and an end time between 8:00 and 8:15, which is the broadcast time of the program M, is specified.

  Subsequently, the appearance count allocating unit 15 matches each keyword with the video material information stored in the video material information storage unit 127 regarding the specified video file name, and specifies a scene related to each keyword ( S207).

  FIG. 15 is a diagram illustrating a configuration example of video material information. Video material information is generated for each video file. For example, in FIG. 15, (1) is video material information regarding “drama A120-1.mpg”, and (2) is drama A120-2.mpg. This is video material information related to mpg. In step S207, other video files are also specified, but are omitted in the description here.

  Video material information related to a video file includes a material file name, a scene name, metadata, a start time stamp, an end time stamp, and the like for each scene included in the video file.

  The material file name is the file name of the video material file before editing the video file. The scene name is the name of the scene. The metadata is a character string indicating terms that characterize the scene. The start time stamp and the end time stamp are time stamps indicating a range corresponding to the scene in the video material file related to the material file name. That is, a video in a range from the elapsed time indicated by the start time stamp to the elapsed time indicated by the end time stamp, starting from the beginning of the video material file, is used for the scene. For example, in (1), the first five-minute video of “Drama A120-1.mpg” is a video of the video material file “Brand B0122” of 0:02:15 to 0:07:15, followed by 1 The video for the minute is a video of the video material file “Brand C0111” from 0: 1: 30 to 0:02:30, and the video for the next 1 minute and 30 seconds is 0:08 of the video material file “Brand A0115”. It is shown that the video of 35 to 0:10:05 and the video of 0: 1: 30 seconds that follows is a video of 0:00:00 to 0:01:30 of the video material file “train 0116”. ing.

  In step S207, for each keyword, a scene including the keyword in metadata is specified. For example, according to the example of FIG. 15, “Brand A” and “Whisky” are included in the metadata of the third scene of “Drama A120-1.mpg”. “Brand B” includes the metadata of the eleventh scene of “drama A120-1.mpg” and drama A120-2.mpg. mpg ”is included in the metadata of the first scene and the metadata of the second scene. “Takehara City” is included in the metadata of the second scene of “Drama A120-1.mpg”.

  Subsequently, for each keyword, the appearance count assigning unit 15 generates the occurrence of the keyword with respect to the time corresponding to the period from the start time to the end time of the scene specified for the keyword, among the times in the audience rating data. The number of times is assigned (S208).

  The start time and end time of the scene specified for the keyword are the start time of the video file including the scene, the order of the scene in the video file, the length of the scene, and the scene before the scene in the video file. And can be calculated based on the time. For example, the scene specified for “Brand B” in “Drama A120-1.mpg” is the first scene. The length of the scene is 5 minutes. Also, the start time of “Drama A120-1.mpg” is “8:01:15” from FIG. Therefore, the start time of the scene is “8:01:15”, and the end time is “8:06:15”. In this case, among the times included in the audience rating data in FIG. 5, 336 that is the number of appearances of “Brand B” is assigned to each time included in “8:01:15 to 8:06:15”. Is done.

  The scene specified for “Takehara City” in “Drama A120-1.mpg” is the second scene. The length of the first scene is 5 minutes, and the start time of “Drama A120-1.mpg” is “8:01:15”. The length of the scene is 1 minute. Therefore, the start time of the scene is “8:06:15”, and the end time is “8:07:15”. In this case, 166 which is the number of appearances of “Takehara-shi” is given to “8:07” among the times included in the audience rating data of FIG.

  When the number of appearances for a plurality of keywords is given for one time, the plurality of appearances are added together.

  As a result of the execution of step S208, for example, information as shown in FIG. 16 is obtained. FIG. 16 is a diagram illustrating an example of the number of appearances assigned at each time based on a match between a keyword and metadata. FIG. 16 shows the number of appearances assigned to each time shown in FIG. In FIG. 16, keywords that contribute to the allocation of the number of appearances are shown in association with each time. Hereinafter, the number of appearances assigned at each time based on the match between the keyword and the metadata is referred to as “appearance number M”.

  Subsequently, the appearance frequency allocation unit 15 matches the details of the corner of the broadcast log (FIG. 6) corresponding to the broadcast time of the program M with each keyword, and specifies the corner related to each keyword (S209). Specifically, for each keyword, a broadcast log that includes the keyword in detail in a corner and whose start time and end time are within the broadcast time of the program M is specified.

  Subsequently, for each keyword, the number-of-appearances assigning unit 15 compares the time corresponding to the period from the start time to the end time of the corner (broadcast log) specified for the keyword, among the times in the audience rating data. The number of appearances of the keyword is assigned (S210).

  FIG. 17 is a diagram illustrating an example of the number of appearances assigned at each time based on a match between a keyword and a broadcast log. For example, the corner details of the first broadcast log shown in FIG. 6 include the character string “Brand B”. Therefore, 366, which is the number of appearances of “Brand B”, is assigned to the time corresponding to the period from the start time to the end time of the broadcast log. In FIG. 6, the corner details of the third broadcast log include the character string “whiskey”. Therefore, 221 as the number of appearances of “whiskey” is assigned to the time corresponding to the period from the start time to the end time of the broadcast log.

  The number of appearances assigned to each time based on the match between the keyword and the broadcast log is hereinafter referred to as “appearance number L”.

  Subsequently, for each time in the audience rating data, the appearance count assigning unit 15 stores the appearance count M and the appearance count L assigned to the time in the time appearance count storage unit 128 (S211).

  Next, the processing procedure executed by the scene specifying unit 16 will be described. FIG. 18 is a flowchart for explaining an example of a processing procedure of scene specifying processing.

In step S <b> 301, the scene specifying unit 16 reads the audience rating data of the program M from the audience rating data storage unit 121. Subsequently, the scene specifying unit 16 reads the number of appearances M and the number of appearances L for each time from the number of appearances storage unit 128 for each time (S302). Subsequently, the scene specifying unit 16 calculates a correction value (t) based on the following equation for each time (S303).
Correction value (t) = (Number of appearances M at time t + Number of appearances L at time t) × Adjustment coefficient The correction value (t) is a value used for correcting the audience rating at time t. Specifically, in the subsequent step, the audience rating is corrected by adding a correction value (t) to the audience rating at time t. The adjustment coefficient is a coefficient for adjusting the number of appearances to a size suitable for audience rating correction. That is, when (number of appearances M at time t + number of appearances L at time t) is directly added to the audience rating, the value obtained after the addition is almost the number of appearances, and the influence of the audience rating is lost. Therefore, an adjustment coefficient is introduced in order to adjust the size of the number of appearances so that the influence of the audience rating is not lost. For example, the adjustment coefficient may be obtained so that the maximum number of appearances is about half of the maximum audience rating, or the adjustment coefficient may be calculated by other methods.

  FIG. 19 is a diagram illustrating an example of calculation results of correction values for each time. FIG. 19 shows the number of appearances M and the number of appearances L, and the correction values calculated based on the number of appearances M and the number of appearances L for each time. FIG. 19 shows an example in which 0.005 is used as the adjustment coefficient.

  Subsequently, the scene identification unit 16 corrects the audience rating by adding the correction value at that time to the audience rating for each age at each time (S304).

  FIG. 20 is a diagram illustrating an example of corrected audience rating data. FIG. 20 shows a result of adding the correction value of FIG. 19 to the audience rating data of FIG. Hereinafter, the corrected value is referred to as “priority”. The ellipse added to FIG. 20 will be described later.

  Subsequently, the scene specifying unit 16 specifies the time with higher priority by age (S305). For example, times up to the highest Nth priority are specified by age. The value of N may be set based on the number of scenes used for the digest version program. In FIG. 20, the highest priority for each age is surrounded by an ellipse. For example, if N is 1, “8:02” is specified for teenagers.

  In FIG. 5, the maximum audience rating for each age is surrounded by an ellipse. When FIG. 20 is compared with FIG. 5, the times corresponding to the values surrounded by the ellipses are different from each other. This is because the content of the posted message posted to the SNS system 30 is considered in FIG. As a result, it is possible to identify the scene of the program M that is highly interested in the viewer, including the interest of the viewer that is difficult to grasp only by the audience rating.

  Subsequently, the scene specifying unit 16 specifies a scene corresponding to the time specified in step S305 for each age (S306). For example, the scene corresponding to “8:02” specified for the teenager is the first of “Drama A120-1.mpg” based on the broadcast progress data (FIG. 14) and the video material information (FIG. 15). Scene ("Distillery B"). Scenes are similarly identified for other ages. Further, when the top N times are specified, a maximum of N scenes are specified for each age.

  Subsequently, the scene specifying unit 16 outputs the identification information of the scene specified in step S306 in order of priority for each age (S307). The scene identification information is, for example, information including the file name of the video file including the scene and the scene name of the scene. The identification information may be output by being displayed on the display device 106, may be output by being printed on a printer, or may be output by another method.

  A person who produces a digest version program of the program M produces a digest version program according to the age with reference to the output information. As a result, it is possible to produce a digest version program in which the scene is specified based not only on the digest version program in which the scene is specified based on the audience rating but also on the content posted to the SNS system 30. Video data of a digest version program created by age is stored in the digest storage unit 129.

  Next, a processing procedure executed when a digest version program is distributed to a certain user will be described. Hereinafter, the certain user is referred to as “user U”.

  FIG. 21 is a flowchart for explaining an example of a processing procedure executed regarding distribution of a digest version program. In FIG. 21, it is assumed that the user U is watching an Internet-connected television.

  In step S <b> 401, the viewing history acquisition unit 17 acquires viewing history data of the Internet-connected television for a certain period of the user U from the viewing information providing apparatus 20 and stores the viewing history data in the viewing history storage unit 130. . For example, viewing history data at the broadcast time of the program M in the vicinity of January 23 when the program M was broadcast is acquired. It is assumed that the program M is a program that is broadcast every day at the same time.

  FIG. 22 is a diagram illustrating a configuration example of viewing history data. In FIG. 22, the channel number of the channel that the user U was watching in the time zone of 8:00 to 8:16 on January 21, January 22, and January 23 is 1 minute. Shown in ticks. For example, at 8:00 on January 21st, channel 101 was viewed, at 8:01, channel 201 was viewed, and at 8:03, channel 101 was viewed.

  Subsequently, the age estimation unit 18 calculates the number of times of viewing of the 101 channel, which is the channel of the program M, for each time of the viewing history data (S402).

  FIG. 23 is a diagram illustrating an example of a calculation result of the number of viewing times. For example, according to FIG. 22, the channel that the user U was watching at 8:00 in any of the three days from January 21 to 23 is the 101 channel. Therefore, in FIG. 23, the number of views for 8:00 is 3.

  Subsequently, the age estimation unit 18 compares the change in the number of viewing times in the time series shown in FIG. 23 with the change in the time series audience rating by age (FIG. 5), and The age is estimated (S403). For example, the age estimation unit 18 estimates the age related to the transition similar to the transition of the number of viewings among the transition of the audience rating by age as the age of the user U. The curve of the graph where the vertical axis is the number of views and the horizontal axis is the time is compared with the curve of the graph according to the age where the vertical axis is the audience rating and the horizontal axis is the time. The most approximate age may be estimated as the age of the user U.

  Subsequently, the digest distribution unit 19 displays a banner (advertisement) of the digest version program of the program M associated with the estimated age and stored in the digest storage unit 129 on the terminal of the user U (S404). . Subsequently, the digest distribution unit 19 distributes the digest version program to the terminal of the user U (S405). As a result, it is possible to increase the possibility that a digest version program that matches the age of the user U is distributed to the user U.

  In the present embodiment, the service of the SNS system 30a is an example of a first information processing service. The service of the SNS system 30b is an example of a second information processing service. The posted message storage unit 31a is an example of a first storage unit. The posted message storage unit 31b is an example of a second storage unit. The video material information storage unit 127 is an example of a third storage unit. The keyword extraction unit 14 is an example of a first specifying unit. The appearance number assigning unit 15 and the scene specifying unit 16 are examples of a second specifying unit.

  As mentioned above, although the Example of this invention was explained in full detail, this invention is not limited to such specific embodiment, In the range of the summary of this invention described in the claim, various deformation | transformation・ Change is possible.

Regarding the above description, the following items are further disclosed.
(Appendix 1)
A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service Identify
Refer to a second storage unit that stores a post message of the second information processing service, which is less immediate than the first information processing service, and a longer than the first period from the broadcast timing of the broadcast program. A keyword related to the broadcast program posted to the second information processing service within a period of 2,
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second Based on a keyword posted to an information processing service, a scene associated with at least one of the identified keywords is identified from the broadcast program.
A broadcast program scene identification method, wherein a computer executes the process.
(Appendix 2)
The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the number of appearances of the keyword associated with each of the scenes in the posted message, and a scene having a relatively large weight is specified.
The scene specifying method according to supplementary note 1, wherein:
(Appendix 3)
The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the audience rating corresponding to the broadcast time of the scene specified based on information indicating the audience rating of the broadcast program in time series, and the number of appearances. Identify relatively large scenes,
The scene specifying method according to supplementary note 2, characterized by:
(Appendix 4)
When a keyword corresponding to a specific scene included in the timetable information of a broadcast program is detected from a posted message to the second information processing service when it is detected from the posted message to the first information processing service On the other hand, an evaluation algorithm is applied in which the decrease in weight according to the increase in the delay period from the broadcast timing of the specific scene to the message posting timing becomes slower.
An evaluation method, wherein a computer executes a process.
(Appendix 5)
The evaluation method according to appendix 4, wherein the first information processing service has a longer message length than that of the second information processing service.
(Appendix 6)
For each of multiple information processing services, accept classification settings according to the immediacy of message posting,
When a keyword corresponding to a specific scene included in the time table information of the broadcast program is detected from a posted message to the information processing service classified into the first classification in which the immediacy of message posting is relatively low, The delay period from the broadcast timing of the specific scene to the message posting timing in the case where the message posting is extracted from the posted message to the information processing service classified into the second classification with relatively high immediacy of message posting Apply an evaluation algorithm that slows down the weighting with increasing
An evaluation method, wherein a computer executes processing.
(Appendix 7)
A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service And the second storage unit in which the post message of the second information processing service that is less immediate than the first information processing service is stored, and the first timing is determined from the broadcast timing of the broadcast program. A first specifying unit for specifying a keyword related to the broadcast program posted to the second information processing service within a second period longer than the period;
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second A second specifying unit that specifies a scene associated with at least one of the specified keywords of the broadcast program based on a keyword posted to the information processing service;
An apparatus for specifying a scene of a broadcast program, comprising:
(Appendix 8)
The second identification unit is associated with a scene associated with a keyword posted to the identified first information processing service and a keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the number of appearances of the keyword associated with each of the scenes in the posted message, and a scene having a relatively large weight is specified.
The scene specifying device according to appendix 7, wherein
(Appendix 9)
The second identification unit is associated with a scene associated with a keyword posted to the identified first information processing service and a keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the audience rating corresponding to the broadcast time of the scene specified based on information indicating the audience rating of the broadcast program in time series, and the number of appearances. Identify relatively large scenes,
The scene specifying device according to appendix 8, characterized in that:
(Appendix 10)
A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service Identify
Refer to a second storage unit that stores a post message of the second information processing service, which is less immediate than the first information processing service, and a longer than the first period from the broadcast timing of the broadcast program. A keyword related to the broadcast program posted to the second information processing service within a period of 2,
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second Based on a keyword posted to an information processing service, a scene associated with at least one of the identified keywords is identified from the broadcast program.
A program that causes a computer to execute processing.
(Appendix 11)
The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the number of appearances of the keyword associated with each of the scenes in the posted message, and a scene having a relatively large weight is specified.
The program according to supplementary note 10, characterized by:
(Appendix 12)
The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the audience rating corresponding to the broadcast time of the scene specified based on information indicating the audience rating of the broadcast program in time series, and the number of appearances. Identify relatively large scenes,
The program according to appendix 11, which is characterized by the above.

DESCRIPTION OF SYMBOLS 10 Scene identification apparatus 11 Viewing information acquisition part 12 SNS setting reception part 13 Post message acquisition part 14 Keyword extraction part 15 Appearance frequency allocation part 16 Scene identification part 17 Viewing history acquisition part 18 Age estimation part 19 Digest distribution part 20 Viewing information provision apparatus 30a, 30b SNS system 31a, 31b Posted message storage unit 100 Drive device 101 Recording medium 102 Auxiliary storage device 103 Memory device 104 CPU
105 interface device 121 audience rating data storage unit 122 broadcast log storage unit 123 SNS setting storage unit 124a, 124b acquisition message storage unit 125 keyword storage unit 126 broadcast progress data storage unit 127 video material information storage unit 128 appearance count storage unit 129 Digest storage unit 130 Viewing history storage unit B bus

Claims (8)

A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service Identify
Refer to a second storage unit that stores a post message of the second information processing service, which is less immediate than the first information processing service, and a longer than the first period from the broadcast timing of the broadcast program. A keyword related to the broadcast program posted to the second information processing service within a period of 2,
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second Based on a keyword posted to an information processing service, a scene associated with at least one of the identified keywords is identified from the broadcast program.
A broadcast program scene identification method, wherein a computer executes the process. The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the number of appearances of the keyword associated with each of the scenes in the posted message, and a scene having a relatively large weight is specified.
The scene specifying method according to claim 1, wherein: The process for identifying the scene is associated with the scene associated with the keyword posted to the identified first information processing service and the keyword posted to the identified second information processing service. For each of the scenes, weighting is performed according to the audience rating corresponding to the broadcast time of the scene specified based on information indicating the audience rating of the broadcast program in time series, and the number of appearances. Identify relatively large scenes,
The scene specifying method according to claim 2, wherein: When a keyword corresponding to a specific scene included in the timetable information of a broadcast program is detected from a posted message to the second information processing service when it is detected from the posted message to the first information processing service On the other hand, an evaluation algorithm is applied in which the decrease in weight according to the increase in the delay period from the broadcast timing of the specific scene to the message posting timing becomes slower.
An evaluation method, wherein a computer executes a process.   The evaluation method according to claim 4, wherein the first information processing service has a longer message length that can be posted than the second information processing service. For each of multiple information processing services, accept classification settings according to the immediacy of message posting,
When a keyword corresponding to a specific scene included in the time table information of the broadcast program is detected from a posted message to the information processing service classified into the first classification in which the immediacy of message posting is relatively low, The delay period from the broadcast timing of the specific scene to the message posting timing in the case where the message posting is extracted from the posted message to the information processing service classified into the second classification with relatively high immediacy of message posting Apply an evaluation algorithm that slows down the weighting with increasing
An evaluation method, wherein a computer executes a process. A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service And the second storage unit in which the post message of the second information processing service that is less immediate than the first information processing service is stored, and the first timing is determined from the broadcast timing of the broadcast program. A first specifying unit for specifying a keyword related to the broadcast program posted to the second information processing service within a second period longer than the period;
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second A second specifying unit that specifies a scene associated with at least one of the specified keywords of the broadcast program based on a keyword posted to the information processing service;
An apparatus for specifying a scene of a broadcast program, comprising: A keyword related to the broadcast program posted to the first information processing service within a first period from the broadcast timing of the broadcast program with reference to the first storage unit storing the posted message of the first information processing service Identify
Refer to a second storage unit that stores a post message of the second information processing service, which is less immediate than the first information processing service, and a longer than the first period from the broadcast timing of the broadcast program. A keyword related to the broadcast program posted to the second information processing service within a period of 2,
The third storage unit in which each scene of the broadcast program and a specific character string are stored in association with each other is referred to, and the keyword posted to the specified first information processing service and the specified second Based on a keyword posted to an information processing service, a scene associated with at least one of the identified keywords is identified from the broadcast program.
A program that causes a computer to execute processing.

Images