JP5017205B2 - Program analysis apparatus, program analysis method, and program analysis program - Google Patents

Description

  The present invention relates to a program analysis apparatus, a program analysis method, and a program analysis program, and in particular, a program analysis apparatus, a program analysis method, and a program analysis that can reduce the burden on the user when specifying the reason why the audience rating has increased. Regarding the program.

  Conventionally, when calculating the audience rating of a TV program, the audience rating survey terminal is installed in a specific household extracted by the audience rating survey company to collect audience information, and the audience rating is calculated based on the collected audience information (For example, refer to Patent Document 1). The calculated audience rating is used as an element for the broadcast station, the advertising company, and the advertiser to determine the number of CM (commercial) viewing households and the advertising fee.

  In addition, broadcaster program producers refer to TV program videos during times of high audience rating in order to improve the quality of the television program and allow more viewers to view the program. Is used to make TV programs by determining the reason for the increase (for example, the moment a professional baseball player hits a home run while a popular actor appears).

JP 2008-92526 A

  However, in the above-described conventional technique, when the reason for the high audience rating is specified, the program producer needs to refer to the TV program scene at the time when the audience rating has increased for each program. There was a problem that the burden on the program producer was large.

  The present invention was made to solve the above-described problems caused by the prior art, and a program analysis apparatus and a program analysis method capable of reducing the burden on the user when specifying the reason why the audience rating has increased. And it aims at providing a program analysis program.

In order to solve the above problems, a program analysis apparatus according to the present invention acquires video information including video and audio information of a television program, analyzes the video included in the acquired video information, and features the video A feature extraction means for extracting the amount, an audience rating change calculating means for obtaining information on the audience rating of a television program that changes over time, and calculating a rate of change in the audience rating; and a feature quantity of the video Change time specifying means for determining whether or not the feature amount of the video has changed, and specifying the time when the feature amount of the video has changed as a feature amount change time; and a rate of change in the audience rating equal to or higher than a first threshold or Extraction target time specifying means for specifying, as an extraction target time zone, a time zone sandwiched between the feature amount change times existing before and after a time that is less than a threshold value of 2, and subtitle text displayed in accordance with the video of the television program The Subtitle text extraction means for extracting subtitle text in the extraction target time zone from the video information; and extracting audio information in the extraction target time zone from the video information and converting the audio information into text. Synthesized by the text synthesizing means, the text synthesizing means for synthesizing the subtitle text extracted by the subtitle text extracting means, and the text converted by the voice text converting processing means; characterized in that a change reason extraction means you extracted as a change reason of the audience rating text.

In the program analysis apparatus according to the present invention as set forth in the invention described above, the reason for change extraction unit extracts a name of a person appearing in a video in the extraction target time zone as the reason for change in the audience rating. characterized in that it.

In the program analysis device according to the present invention, in the above invention, the change reason extraction unit compares the extracted reasons for change of the audience rating, and determines that the overlapping change reason is the final change reason. Features.

In the program analysis method according to the present invention, the program analysis apparatus acquires video information including video and audio information of a television program , analyzes the video included in the acquired video information, and analyzes the feature amount of the video. A feature extraction step for extracting a TV program, an audience rating change calculating step for obtaining a rating ratio of a change in the audience rating, and information on a rating change of the TV program that changes with the passage of time. A change time specifying step of determining whether or not the feature amount of the video has changed as a feature amount change time, and a rate of change in the audience rating equal to or higher than a first threshold or a second An extraction target time specifying step for specifying a time zone sandwiched between the feature amount change times existing before and after a time that is less than a threshold value as an extraction target time zone, and characters displayed in accordance with the video of the television program A subtitle text extraction step for extracting subtitle text in the extraction target time zone from the video information including text, and extracting audio information in the extraction target time zone from the video information, and converting the audio information into text A voice text processing step to be converted; a text synthesis step for synthesizing the subtitle text extracted by the subtitle text extraction step; and the text converted by the voice text processing step; and a text synthesis step. And a change reason extracting step of extracting the text as a reason for change in the audience rating .

In addition, the program analysis program according to the present invention acquires, from a computer, video information including video and audio information of a TV program , analyzes the video included in the acquired video information, and extracts the feature amount of the video A feature extraction procedure, a viewing rate change calculation procedure for obtaining information on the audience rating of a TV program that changes over time, and calculating a rate of change in the audience rating, and whether the feature amount of the video has changed A change time specifying procedure for determining whether or not the feature amount of the video has changed as a feature amount change time, and a rate of change in the audience rating is greater than or equal to a first threshold or a second threshold the extraction target time specifying step for specifying a time zone as the extraction target time zone sandwiched feature quantity change time, subtitle text displayed in accordance with the video of the television program that exists before and after the time that is less than A subtitle text extraction procedure for extracting subtitle text in the extraction target time zone from the video information, and extracting audio information in the extraction target time zone from the video information, and converting the audio information to text Synthesized by the text synthesis procedure, the text synthesis procedure for synthesizing the subtitle text extracted by the subtitle text extraction procedure, and the text converted by the speech text processing procedure And a change reason extraction procedure for extracting text as a change reason for the audience rating .

According to the present invention, it is possible to reduce the user's burden, to identify the reason for viewing聴率is increased.

First, an outline of processing of the program analysis apparatus according to the present embodiment will be described. The program analysis apparatus according to the present embodiment acquires video information including video and audio information of a TV program and information on the audience rating of a TV program that changes over time, and acquires the acquired TV program video Of the information, the video information of the portion where the rate of change in the audience rating is equal to or higher than the first threshold or lower than the second threshold is stored in the storage device. Then, the program analysis device extracts a video feature (for example, an appearing actor) or a voice feature (for example, a voice called “home run” repeatedly called by a commentator) included in the video information stored in the storage device. The user is notified of the extracted information as the reason why the audience rating has increased.

As described above, the program analyzing apparatus according to the present embodiment stores the video information of the portion in which the rate of change in the audience rating is equal to or higher than the first threshold value or lower than the second threshold value in the storage device, and is stored in the storage device. Because the video information and audio characteristics are notified to the user as a factor of high audience rating from the recorded video information, the user needs to refer to each TV program and audience rating to investigate the reason why the audience rating has increased. The burden on the user can be reduced.

Next, the configuration (example) of the system according to the present embodiment will be described. FIG. 1 is a diagram illustrating a configuration of a system according to the present embodiment. As shown in the figure, this system includes a broadcast station terminal 30, a client terminal 50, and a program analysis apparatus 100, which are connected to each other by a network 10. Here, for convenience of explanation, only the broadcasting station terminal 30 and the client terminal 50 are shown, but this system is assumed to further include a broadcasting station terminal and a client terminal.

Among these, the broadcast station terminal 30 is a device that transmits various types of information related to the television program (for example, video data and caption data of the television program) to the program analysis device 100. The client terminal 50 is a device that displays information on a television program distributed from the program analysis device 100 using a widget function.

The program analysis apparatus 100 uses the widget function to distribute TV program information and the like to the client terminal 50, stores video data of a portion where the rate of change in the audience rating is equal to or greater than a predetermined value, and stores the stored video It is a device that extracts factors that have obtained high audience ratings from data.

  Next, a configuration (example) of the client terminal 50 illustrated in FIG. 1 will be described. FIG. 2 is a functional block diagram illustrating the configuration of the client terminal 50 according to the present embodiment. As shown in the figure, the client terminal 50 includes an input unit 51, an output unit 52, a communication control IF unit 53, an input / output control IF unit 54, a storage unit 55, and a control unit 56.

  Among these, the input unit 51 is an input means for inputting various types of information (for example, a channel of a TV program to be viewed), and includes a keyboard, a mouse, a microphone, and the like. A monitor (output unit 52) described later also realizes a pointing device function in cooperation with the mouse.

  The output unit 52 is a unit that outputs various types of information (for example, video and audio of a television program), and is configured by a monitor (or display, touch panel), a speaker, and the like.

The communication control IF unit 53 is a unit that mainly controls communication with the program analysis device 100. The input / output control IF unit 54 is means for controlling input / output of data by the input unit 51, the output unit 52, the communication control IF unit 53, the storage unit 55, and the control unit 56.

  The storage unit 55 is a storage unit that stores data and programs necessary for various processes performed by the control unit 56.

  The control unit 56 has an internal memory for storing programs defining various processing procedures and control data, and is a control means for executing various processes using these. In particular, as closely related to the present invention, as shown in FIG. 2, a widget control unit 56a is provided.

  Here, the widget control unit 56a uses the widget function to monitor various information (calendar, time, weather forecast, TV program, etc .; in particular, the TV program in this embodiment) on the screen of the monitor (output unit 52). It is a means to display. In the following description, an area on the screen where the widget control unit 56a displays various types of information is referred to as a widget window.

Widget control module 56a, a user identification number via the input unit 51, when receiving the password, and the user identification number, the program analyzer service initiation request by sending the password to the program analyzer 100 100 To do.

After the service from the program analysis device 100 is started, the widget control unit 56a acquires the channel information of the TV program that the user desires to view via the input unit 51, and the channel analysis information is displayed on the program analysis device. The information of the TV program corresponding to the channel information is acquired from the program analysis apparatus 100.

The TV program information transmitted from the program analysis apparatus 100 includes video data of the TV program, caption data, and the like. When acquiring the TV program information, the widget control unit 56a displays the acquired TV program information on the widget window.

Next, a configuration (example) of the program analysis apparatus 100 shown in FIG. 3 will be described. FIG. 3 is a functional block diagram of the configuration of the program analysis apparatus 100 according to the present embodiment. As shown in the figure, the program analysis apparatus 100 includes an input unit 110, an output unit 120, a communication control IF unit 130, an input / output control IF unit 140, a storage unit 150, and a control unit 160. .

  Among these, the input unit 110 is an input unit that inputs various types of information, and includes a keyboard, a mouse, a microphone, and the like. A monitor (output unit 120) described later also realizes a pointing device function in cooperation with the mouse.

  The output unit 120 is an output unit that outputs various types of information, and includes a monitor (or display, touch panel), a speaker, and the like.

  The communication control IF unit 130 is a unit that mainly controls communication between the broadcast station terminal 30 and the client terminal 50. The input / output control IF unit 140 is a unit that controls input / output of data by the input unit 110, the output unit 120, the communication control IF unit 130, the storage unit 150, and the control unit 160.

  The storage unit 150 is a storage unit that stores data and programs necessary for various processes performed by the control unit 160. As shown in FIG. 3, the storage unit 150 includes a program management table 150a, a viewer management table 150b, a burst determination table 150c, and a video data management table 150d.

  Among these, the program management table 150 a is a table for managing information on each television program distributed to the client terminal 50. FIG. 4 is a diagram illustrating an example of a data structure of the program management table 150a according to the present embodiment. As shown in FIG. 4, the program management table 150a stores a program identification number, a program name, a genre, a channel, broadcast date and time, program video data, and caption data in association with each other. Here, the program video data is information including video and audio of the television program, and the caption data is information of caption displayed in accordance with the video of the television program.

  The viewer management table 150b is a table for managing the TV program being viewed by the user. FIG. 5 is a diagram illustrating an example of a data structure of the viewer management table 150b according to the present embodiment. As shown in the figure, this viewer management table 150b stores a user identification number and a program identification number in association with each other. In the example shown in the first row of FIG. 5, information indicating that the user with the user identification number “2001” is watching the television program with the program identification number “1002” is stored.

The burst determination table 150c is a table for managing the instantaneous audience rating of each television program, the reason why the audience rating has changed, and the like. FIG. 6 is a diagram illustrating an example of a data structure of the burst determination table 150c according to the present embodiment. As shown in FIG. 6, this burst determination table 150c stores a program identification number, a date and time, an instantaneous audience rating, a viewing change rate, a change reason, and moving image storage location information in association with each other.

  Of these, the date and time indicates the date and time when the viewing change rate (for example, the amount of change in the viewing rate per unit time) is equal to or greater than a predetermined value, and the instantaneous audience rate is equal to or greater than the predetermined value. Indicates the audience rating at the time.

The reason for change indicates the reason why the audience rating (viewing rate change) has changed. The change reason information is created by a change reason extraction unit 160e described later. The moving image storage location information is information for specifying a position for storing information (video, audio, caption) of the television program before and after the audience rating has changed.

  The video data management table 150d is a table for managing video data (including audio) and subtitle data of a portion where the audience rating has changed. FIG. 7 is a diagram illustrating an example of the data structure of the video data management table according to the present embodiment. As shown in FIG. 7, the video data management table 150d stores an address, a program identification number, video data, and audio text data in association with each other.

  Of these, the address is an address for identifying a moving image storage location, and the video data is video data of a television program at a portion where the audience rating has changed. The audio text data is audio text data in the video data of the video data management table 150d.

  Returning to the description of FIG. 3, the control unit 160 is a control unit that has an internal memory for storing programs and control data that define various processing procedures, and executes various processes using these. Particularly, as closely related to the present invention, as shown in FIG. 3, a program distribution processing unit 160a, a viewing rate calculation unit 160b, a viewing change rate calculation unit 160c, a video analysis processing unit 160d, A reason extraction unit 160e.

  Among these, the program distribution processing unit 160 a is means for distributing information on the television program to the client terminal 50. First, the program distribution processing unit 160a receives a user identification number and a password from the client terminal 50, and starts a service for the client terminal 50 when it is determined that the combination of the user identification number and the password is appropriate. To do.

  When the channel distribution information is acquired from the client terminal 50 after starting the service for the client terminal 50, the program distribution processing unit 160a extracts the information of the TV program corresponding to the channel information from the program management table 150a and extracts it. The TV program information is transmitted to the client terminal 50.

  Then, the program distribution processing unit 160a registers the user identification number and the program identification information corresponding to the television program being transmitted in association with the user identification number in the viewer management table 150b. In addition, when acquiring information indicating that the channel has been switched from the client terminal 50, the program distribution processing unit 160a changes the combination of the user identification number and the program identification number, and updates the viewer management table 150b. .

  On the other hand, when receiving various information related to the television program from the broadcast station terminal 30, the program distribution processing unit 160a registers the various information related to the received television program in the program management table 150a.

  The audience rating calculation unit 160b is means for calculating an audience rating for each TV program with reference to the viewer management table 150b. For example, when the audience rating calculating means calculates the audience rating of the television program “A”, the viewers who are viewing the television program “A” with respect to the total number of viewers using each television program. The audience rating is calculated by calculating the ratio of numbers. The audience rating calculator 160b calculates the audience rating of the television program every predetermined time (or every unit time), and outputs the calculation result to the audience change rate calculator 160c.

  The audience change rate calculation unit 160c acquires the audience rating calculation result from the audience rating calculation unit 160b, and calculates the audience change rate (for example, the rate of change in audience rate per unit time) based on the obtained calculation result. It is means to do. The viewing change rate calculation unit 160c outputs the calculated viewing change rate information and the program identification number of the corresponding television program to the video analysis processing unit 160d.

  The video analysis processing unit 160d is means for extracting information on the TV program (video data and audio text data shown in FIG. 7) in a portion where the viewing change rate is equal to or higher than the first threshold value or lower than the second threshold value. FIG. 8 is a functional block diagram of the configuration of the video analysis processing unit 160d according to the present embodiment. As shown in FIG. 8, the video analysis processing unit 160d includes an image feature amount extraction unit 161, a feature amount change time extraction unit 162, a subtitle text extraction unit 163, an audio text conversion processing unit 164, and a text synthesis process. Unit 165 and a television information registration unit 166.

The image feature amount extraction unit 161 is a unit that analyzes an image (video) of a television program that changes with time and extracts the feature amount of the analyzed image. For example, the image feature amount extraction unit 161 divides an image of a television program into a plurality of blocks, and calculates an average density (or luminance, saturation RGB “Red Green Blue” value, etc .; the same applies hereinafter) in each block. Are extracted as feature quantities. The image feature amount extraction unit 161 outputs the extracted feature amount information of the image and the time information corresponding to the image (the time of the image in the video data) to the feature amount change time extraction unit 162.

  The feature amount change time extraction unit 162 acquires the feature amount information of each image from the image feature amount extraction unit 161, determines whether or not the feature amount of the image has changed, and when the feature amount of the image has changed Is a means for specifying the time when the feature amount of the image changes.

  For example, the feature amount change time extraction unit 162 compares the average density of each block of the image with the average density of each block of the previous image, and as a result of the comparison, the difference in average density is equal to or greater than a predetermined value. The number of blocks is calculated. Then, the feature amount change time extraction unit 162 determines that the feature amount of the image has changed when the number of blocks in which the difference in average density exceeds a predetermined value exceeds the specified number.

  When the feature amount change time extraction unit 162 determines that the feature amount of the image has changed, the time associated with the image is used as the time when the feature amount of the image has changed (hereinafter, feature amount change time). Extract. The feature amount change time extraction unit 162 outputs the extracted feature amount change time information to the caption text extraction unit 163, the voice text conversion processing unit 164, and the television information registration unit 166.

  The subtitle text extraction unit 163 is means for extracting subtitle data (subtitle text) of a television program in a portion where the viewing change rate is equal to or higher than the first threshold value or lower than the second threshold value. Specifically, the subtitle text extraction unit 163 first sets a time when the rate of change in the audience rating is equal to or higher than the first threshold or lower than the second threshold as a reference time, and changes in feature quantities existing before the reference time. The time and the feature amount change time existing after the reference time are determined.

  Then, the caption text extraction unit 163 extracts, from the program management table 150a, the text of the caption data in the time period sandwiched between the determined feature amount change times. In the following description, a time zone sandwiched between feature amount change times existing before and after the reference time is referred to as an extraction target time.

  The voice text processing unit 164 is means for extracting voice data included in the extraction target time from the program management table 150a and converting the extracted voice data into text data. Note that any known method may be used as a method for converting voice data into text data. For example, by preparing a table in which each character is associated with an audio frequency and comparing the table with audio data in the program, the audio data can be converted into text data.

  The text synthesis processing unit 165 is a unit that synthesizes the subtitle text extracted by the subtitle text extraction unit 163 and the text converted by the audio text conversion processing unit 164. Then, the text synthesis processing unit 165 registers the synthesized text information in the audio text data of the video data management table 150d in association with the corresponding program identification number.

  The TV information registration unit 166 is means for extracting video data (video data of the corresponding TV program) existing at the extraction target time from the program management table 150a. The television information registration unit 166 registers the extracted video data in the video data management table 150d in association with the corresponding program identification information. The television information registration unit 166 registers the addresses corresponding to the registered video data and audio text data in the moving image storage location information (see the burst determination table 150c in FIG. 6).

  The change reason extraction unit 160e is a means for extracting the reason why the audience rating has changed based on the video data management table 150d. First, the video data management table 150d extracts the video data of the video data management table 150d, and specifies, for example, the names of actors (actor names) appearing in the video data as video characteristics.

  Any known technique may be used as the method by which the change reason extraction unit 160e identifies the actor name appearing in the video data. For example, the change reason extraction unit 160e holds in advance a table in which an actor name and actor characteristics (for example, eyes, nose, mouth size, eye interval, etc.) are associated with each other, and the table and video. The actor name is specified by comparing with the data.

  The change reason extraction unit 160e registers the characteristic (for example, actor name) of the specified video data in the burst determination table 150c as the reason for change in association with the program identification number. Also, the change reason extraction unit 160e registers the audio text data (audio characteristics) of the video data management table 150d in the burst determination table 150c as the change reason.

Next, a processing procedure of the program analysis apparatus 100 according to the present embodiment will be described. FIG. 9 is a flowchart illustrating the processing procedure of the program analysis apparatus 100 according to the present embodiment. As shown in FIG. 9, the program analysis apparatus 100 receives a service request from the client terminal 50 (step S101), and the program distribution processing unit 160a distributes a television program (step S102).

  Then, the audience rating calculation unit 160b calculates the audience rating (step S103), and the video analysis processing unit 160d determines whether or not the audience change rate is equal to or higher than the first threshold or lower than the second threshold (step S104). ). If the viewing change rate is less than the first threshold and greater than or equal to the second threshold (step S105, No), the process proceeds to step S108.

On the other hand, when the viewing change rate is greater than or equal to the first threshold or less than the second threshold (step S105, Yes), the video analysis processing unit 160d has the viewing change rate greater than or equal to the first threshold or less than the second threshold. The video data and audio text data of the part that has become are stored (step S106).

Then, the change reason extraction unit 160e extracts the change reason from the stored video data and audio text data (step S107). The program distribution processing unit 160a determines whether or not the distribution of the television program has been completed (step S108). If the distribution has not been completed (step S109, No), the distribution of the television program is continued (step S110). The process proceeds to step S103. On the other hand, if the distribution of the television program has been completed (step S109, Yes), the distribution of the television program is stopped (step S111).

As described above, the program analysis apparatus 100 according to the present embodiment stores, in the storage device, video information of a portion where the rate of change in the audience rating is equal to or higher than the first threshold value or less than the second threshold value. The video information and audio characteristics are notified to the user as a factor of high audience rating from the video information stored in the video information, so the user refers to each TV program and audience rating to determine why the audience rating has increased. There is no need to investigate, and the burden on the user can be reduced.

  Note that the change reason extraction unit 160e illustrated in FIG. 3 compares each change reason stored in the burst determination table 150c when extracting the change reason that the audience rating has changed, and finally determines the overlapping change reason. It may be extracted as a reason for change. Thus, the reason for change can be improved by the change reason extraction unit 160e comparing each change reason and narrowing down the reason for change.

By generating a change reason extraction program that describes the change reason extraction method of the program analysis apparatus 100 described in the above embodiment in a computer-executable format, and causing the computer to execute the change reason extraction program, It is also possible to realize the same processing as the program analysis apparatus 100 of the present invention. At this time, the change reason extraction program may be stored in a computer-readable storage medium, and the change reason extraction program may be introduced into the computer using the storage medium.

  The present invention can be applied when it is necessary to extract the reason why the audience rating of a television program has changed.

It is a figure which shows the structure of the system concerning a present Example. It is a functional block diagram which shows the structure of the client terminal concerning a present Example. It is a functional block diagram which shows the structure of the program analyzer concerning a present Example. It is a figure which shows an example of the data structure of the program management table concerning a present Example. It is a figure which shows an example of the data structure of the viewer management table concerning a present Example. It is a figure which shows an example of the data structure of the burst determination table concerning a present Example. It is a figure which shows an example of the data structure of the video data management table concerning a present Example. It is a functional block diagram which shows the structure of the video analysis process part concerning a present Example. It is a flowchart which shows the process sequence of the program analyzer concerning a present Example.

30 broadcast station terminal 50 client terminal 51,110 input unit 52,120 output unit 53,130 communication control IF unit 54,140 input / output control IF unit 55,150 storage unit 56,160 control unit 56a widget control unit 100 program analyzer 150a program management table 150b viewer management table 150c burst determination table 150d video data management table 160a program distribution processing unit 160b audience rating calculation unit 160c viewing change rate calculation unit 160d video analysis processing unit 160e change reason extraction unit

Claims (5)

Feature extraction means for acquiring video information including video and audio information of a television program, analyzing video included in the acquired video information, and extracting a feature amount of the video;
An audience rating change calculating means for obtaining information on the audience rating of a TV program that changes over time, and calculating a rate of change in the audience rating;
A change time specifying means for determining whether or not the feature amount of the video has changed, and specifying a time at which the feature amount of the video has been changed as a feature amount change time;
Extraction target time specifying means for specifying, as an extraction target time zone, a time zone sandwiched between the feature amount change times that exist before and after a time when the rate of change in the audience rating is greater than or equal to the first threshold value or less than the second threshold value When,
Subtitle text extraction means for extracting subtitle text in the extraction target time zone from the video information including subtitle text displayed in accordance with the video of the television program;
Voice text processing means for extracting voice information in the extraction target time zone from the video information and converting the voice information into text;
Text synthesizing means for synthesizing the subtitle text extracted by the subtitle text extracting means and the text converted by the audio text processing means;
Means out changes reasons extraction you extract the combined text as the change reason of the viewing rate by the text combining means,
An apparatus for analyzing a program. The program analysis apparatus according to claim 1, wherein the change reason extraction unit extracts a name of a person who appears in a video in the extraction target time zone from the video information as a change reason for the audience rating. . The program analysis apparatus according to claim 1, wherein the change reason extraction unit compares the extracted reasons for change in the audience rating, and determines an overlapping change reason as a final change reason . Program analyzer
A feature extraction step of acquiring video information including video and audio information of a television program, analyzing a video included in the acquired video information, and extracting a feature amount of the video;
An audience rating change calculating step of obtaining information on the audience rating of a TV program that changes over time, and calculating a rate of change in the audience rating;
A change time specifying step of determining whether or not the feature amount of the video has changed, and specifying a time at which the feature amount of the video has been changed as a feature amount change time;
Extraction target time specifying step of specifying, as an extraction target time zone, a time zone between the feature amount change times that exists before and after the time when the rate of change in the audience rating is equal to or higher than the first threshold value or less than the second threshold value. When,
A subtitle text extraction step of extracting subtitle text in the extraction target time zone from the video information including subtitle text displayed in accordance with the video of the television program;
Extracting audio information in the extraction target time zone from the video information, and converting the audio information into text;
A text synthesis step of synthesizing the subtitle text extracted by the subtitle text extraction step and the text converted by the speech text processing step;
A change reason extraction step for extracting the text synthesized by the text synthesis step as a reason for the change in the audience rating;
A program analysis method characterized by comprising: On the computer,
A feature extraction procedure for acquiring video information including video and audio information of a TV program, analyzing the video included in the acquired video information, and extracting a feature amount of the video;
An audience rating change calculation procedure for obtaining information on the audience rating of a TV program that changes over time, and calculating a rate of change in the audience rating;
A change time specifying procedure for determining whether or not the feature amount of the video has changed, and specifying a time at which the feature amount of the video has been changed as a feature amount change time;
Extraction target time specifying procedure for specifying, as an extraction target time zone, a time zone between the feature amount change times that exists before and after the time when the rate of change in the audience rating is equal to or higher than the first threshold value or less than the second threshold value. When,
A subtitle text extraction procedure for extracting subtitle text in the extraction target time zone from the video information including subtitle text displayed in accordance with the video of the television program;
Extracting voice information in the extraction target time zone from the video information, and converting the voice information into text;
A text synthesis procedure for synthesizing the subtitle text extracted by the subtitle text extraction procedure and the text converted by the speech text processing procedure;
A change reason extraction procedure for extracting the text synthesized by the text synthesis procedure as a change reason for the audience rating;
A program analysis program characterized in that the program is executed.

Images