JP4722116B2 - Method and apparatus for measuring effect of TV CM based on life time survey - Google Patents

Description

  The present invention analyzes a relationship between a broadcast plan of a CM broadcast on TV, a fact that an individual viewer actually views a CM, and an effect of the CM based on a life time survey of a monitor viewer -It relates to a device for evaluation.

  In order for the CM to be effective, it goes through several processes such as CM transmission by broadcasting, contact between the broadcasted CM and the consumer, memory by the consumer, and inducing purchase behavior. Therefore, it is necessary to create a CM drop plan in consideration of trends such as the number of broadcasts, the arrival rate to consumers, the memory rate by consumers, and changes in sales. Usually, as an indicator of the arrival rate to consumers, the number of CM broadcasts for each program of each TV station is recorded, and a total value obtained by multiplying the number of broadcasts and the CM viewing rate estimated from the program viewing rate is used. Further, as a CM storage rate index, CM favorableness tabulated based on a questionnaire is used.

  In order to measure the program audience rating, a dedicated device is installed at the viewer's house, and the characteristics of the viewing channel and broadcast are automatically determined and transmitted, and this is compared with the program database installed at the center to determine the audience rating. A calculation method is used.

  As described above, it has been pointed out that when the CM audience rate is estimated from the program audience rate and used as an indicator of the arrival rate, there is a deviation from the CM memory rate seen from the viewer side. The reason is that many viewers switch channels immediately after the program broadcast and pass the CM, or switch channels one after another during the CM broadcast, and the program viewership rate and the CM viewership rate do not match. Is done. Therefore, it has been desired to directly calculate the audience rating of the CM itself, instead of estimating it from the program audience rating. However, the CM has a difficult problem such as the broadcast schedule is not disclosed in advance and the broadcast time may be changed according to the program content.

Japanese Laid-Open Patent Publication No. 2000-134167 (Patent No. 3056720) is a technique for investigating household audience ratings of TV / radio CMs. A dedicated device is installed in a receiver at a viewer's house to view broadcast CMs. A technique for investigating the CM audience rate by directly counting the fact that the viewer actually viewed the program is not estimated from the program audience rating.
JP 2000-134167 A

Japanese Patent Laid-Open No. 2001-339744 is a technique for recording a broadcast record of a specific television CM, and a technique for automatically recording a CM broadcast record using identification information for each CM added to the broadcast in advance. Is disclosed. However, since this technology adds identification information to CMs in advance, it is difficult to apply this technology to all CMs broadcast on a certain channel.
JP 2001-339744 A

Japanese Patent Application Laid-Open No. 2006-67229 is a technique for automatically collecting and accumulating records of personal viewing of a television / radio CM. In a television receiver of a viewer monitor, ordinary information to which no identification information is added is disclosed. A technique is disclosed in which features are extracted and accumulated from broadcast audio and the like, and this is periodically collated with feature information of each CM stored in the center to record a CM viewing record with matching features. The feature information is collated by changing the amount of movement of the window of interest according to the statistical similarity between the feature information of the input audio or video signal and the feature information of the reference signal. It has also been shown to employ an active search technique that can be quickly determined.
JP 2006-67229 A

  The recording of broadcast recordings and viewing records by dedicated equipment has the advantage that the audience rating can be aggregated immediately after broadcasting, and there is an advantage that less manpower is required for input, etc. However, it is not easy, and has not yet spread as a CM audience rating measurement technique. Furthermore, more complicated devices are required to perform individual audience rating measurement and multi-channel support.

  The survey based on the questionnaire does not require a dedicated terminal device, and has an advantage that other useful information such as an individual audience rating and product preference can be collected simultaneously with the audience rating. However, since the CM is of a nature that can be seen accompanying the program, it is difficult to remain in memory, and it is difficult to collect accurate viewing content and viewing time after the fact. In addition, there is a problem of how viewing data collected by a questionnaire is linked to a specific CM work or product.

  The present invention provides an effect measurement device for a television CM that can index the degree of contact between all CMs and viewers and objectively analyze the effects of CMs based on a questionnaire survey of viewer monitors.

According to the first aspect of the present invention, a broadcast is received, audio or video feature information is extracted, compared with reference information stored in advance in the reference information storage means, and a CM including the matching CM identification information and the broadcast date and time The broadcast history is recorded in the first recording means, the viewing information for each life time frame collected from the viewer monitor is recorded in the second recording means, and for each CM, each life time is recorded from the first recording means. The CM broadcast history is extracted for each frame, the broadcast frequency is calculated, the number of viewer monitors who have watched the TV in the same life time frame is extracted from the second recording means, and the broadcasts in all the life time frames are extracted. The above problem is solved by a method for measuring the effect of a television CM that calculates the average CM encounter degree of the viewer from the product of the frequency and the number of people and stores it in the third recording means together with the CM identification information.

The broadcast CM effect measurement method of the present invention monitors all broadcast channels to be analyzed, records CM broadcast results, records CM input history, and monitors the viewer for each life time frame based on a questionnaire or the like. Records information (viewing information) etc. as to whether or not the broadcast subject to the survey was being viewed. The average CM encounter level of the viewer is calculated for each CM based on the CM broadcast record and viewing information, and the effect of the TV CM is measured.

  In this specification, what is called a life time frame is a time frame obtained by dividing a broadcast time zone of one day of weekdays or holidays into appropriate units (for example, every 15 minutes). It is the unit that can be expressed best and is the standard for CM input. The viewing information is data indicating whether or not the viewer monitor was watching TV at that time, and does not specify the channel or CM, so it can be easily collected by a diary questionnaire and the viewer's lifestyle It seems that it corresponds well to. Therefore, the average audience encounter level calculated using this is a good index for creating a CM insertion plan. As the broadcast frequency used for calculating the average CM encounter degree, the actual number of CM broadcasts for each channel may be used, or an appropriate multiplier may be set for each channel.

  The present invention uses a technique for extracting feature information from a normal broadcast and specifying a CM to create reference information for CM input history, and by combining it with viewing information, a broadcast plan of an arbitrary CM can be obtained through a questionnaire. It was linked with the life time survey.

  According to the second aspect of the present invention, reference information storage means for storing feature information of CM audio or video as original feature information, information input means for inputting audio or video information from a broadcasting device, and input audio or video information Feature extraction means for extracting feature information from the CM, CM detection means for detecting that a specific CM is broadcast by comparing the extracted feature information with the stored original feature information, and the detected CM CM input history creating unit having CM input history recording means for recording CM broadcast history including the identification information and broadcast date and time, input means, CM identification information inputted from the input means, advertiser identification information, target product identification Information, CM work information creation unit having CM work recording means for recording work attribute information, input means, identification information of viewer monitor based on inputted questionnaire result, monitor Viewer information recording unit and CM work recording means having life information recording means for recording the attribute information and the viewing information which is the fact that the viewer monitor has watched the broadcast to be investigated in each life time frame The CM broadcast history is extracted from the CM input history recording means to calculate the broadcast frequency for the CM recorded on the TV, the viewing information is extracted from the life recording means for each life time frame, and the TV is viewed. The number of viewer monitors that have been used is calculated, the product of the broadcast frequency and the number of people is calculated for all time frames, the viewer encounter level of the CM work is calculated, and the calculated viewer encounter level is calculated as the CM identification information, Effect of television CM comprising CM effect evaluation means for recording in CM evaluation recording means together with advertiser identification information and target product identification information, and CM effect evaluation section having output means for outputting evaluation results A constant unit, solving the above problems.

  According to a third aspect of the present invention, the viewer information recording unit, together with the input viewer monitor identification information and attribute information, CM identification information of the CM selected as having a favorable impression by the viewer monitor, each selected CM Favorability recording means for recording likability factor information, which is information on which of the predetermined likability factors, and trial preparation information, which is information indicating whether the CM target product has trial preparation, The CM effect evaluation unit of the CM effect evaluation unit extracts a record of CM selection by the viewer monitor from the preference recording unit for each CM, and the value obtained by dividing the selected number by the total number of viewer monitors, Calculate and calculate the number of viewers with positive values for each positive factor information and the number of intention-to-trial votes with the number of viewer monitors with positive values for trial preparation information. 3. The television CM effect measuring device according to claim 2, wherein the preference, the factor-specific preference and the number of trial intention votes are recorded together with the CM identification information, the attribute information, and the audience encounter level in the CM evaluation recording means. Solve the above problems.

  Together with the survey of viewing information, by collecting and recording information on the TV CM memory rate and CM merchandise sensitivity, the relationship between the CM rate calculated for each CM and the CM memory rate and favorability Added verification.

  According to a fourth aspect of the present invention, whether or not the life information recording means of the viewer information recording unit is at home in each life time frame together with identification information, attribute information, and viewing information of the viewer monitor. 4. The television CM effect measuring device according to claim 2, further comprising: home information, which is a fact of the above, and viewing attitude information, which is a fact of whether or not each TV is dedicated to watching TV in each life time frame. Thus, the above problem is solved.

By the method or apparatus according to the present invention, the CM encounter information, the CM identification information, the likability, and the factors of the CM works are associated with the CM identification information, the advertiser identification information, and the target product identification information. It is possible to create a recording medium that records the favorable sensitivity for each factor and the number of trial intention votes .

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

  FIG. 1 shows an example of the device configuration of a television CM effect measuring apparatus according to the present invention. This apparatus is connected to each other via a premises wiring, and can be communicated by a known communication means. DB1 and CB2 are provided. Further, auxiliary input / output means can be further provided.

  The CM insertion history creation device 1 is installed in a place where the audio and video of a television broadcast can be detected, and detects the CM broadcast from the received broadcast audio or video while referring to the reference information database in the DB 1. Is recorded in the CM input history database in DB2, and one is installed for each broadcast channel to be investigated. The CM work information recording device 2 is a device for recording CM work information input from a keyboard or the like in a CM work database in the DB 2. The viewer information recording device 3 is for recording the life information of the viewer monitor input from the keyboard, the optical reader, or the like and the preference information for the product in the life time survey database and the CM preference database in the DB 2, respectively. Device. The CM evaluation apparatus 4 is an apparatus for analyzing and evaluating the effect of each CM based on another database and recording it in a CM evaluation database in the DB 2 in a format that allows the evaluation to be called as needed.

  The CM input history creation device 1 includes a reference information database 1a that stores audio or video feature information of a television CM as original feature information, an information collection unit 1b that collects audio or video information from a television broadcast receiver, and a collection CM extraction unit for extracting feature information from the extracted audio or video information, and CM detection for detecting that a specific CM is broadcast by comparing the extracted feature information with the stored original feature information Means 1d and a CM insertion history database 1e for recording the detected CM identification information together with the broadcast time information.

  The reference information database 1a stores the original sound feature information extracted from the original sound of the television CM along with the CM identification information. The information input unit 1b acquires digital audio data of audio or video broadcast by a known unit such as a microphone or an A / D converter. The feature extraction unit 1c extracts feature information from the time-series digital audio data output from the information input unit 1b. As a feature extraction method, for example, a method described in JP-A-2002-41569 may be used. The CM detection means 1d reads out identification information and original sound feature information of all CMs that may be broadcast from the reference information database 1a, stores them in the built-in memory, and sequentially compares the extracted feature information with the original sound feature information. Thus, the CM being broadcast is specified. Then, the identification information and date / time information of the specified CM are recorded in the CM insertion history database 1e in the DB2.

  For the comparison and collation of the feature information by the CM detection means 1d, the same kind of active search technique described in Japanese Patent Application Laid-Open No. 2000-31343 can be used. CM viewing survey devices using this technology have been commercialized by Nippon Telegraph and Telephone Corporation. Broadcast recordings created using this technology may have omissions due to changes in the CM broadcast format, etc., but daily corrections can be made by referring to the CM work database using an auxiliary input device as appropriate. Can be converted into data on the same day.

  The CM work information recording device 2 records the CM work information input from the input means 2a such as a keyboard in the CM work database 2b. The CM work information includes the work code, the advertiser code, the target product code, the advertiser's industry information, the work age (the number of months since the work started broadcasting), the number of aired seconds, and other attribute information. . In addition, since a plurality of commercials may be produced in accordance with the broadcast format or the like for one commercial work, it includes commercial identification information attached separately to all broadcast forms. The CM work information is basically manually input every time a specialist who is constantly watching TV discovers a new broadcast format of a new CM or a registered CM. Also, feature information is registered in the reference information database 1a, and CM identification information and date / time information are recorded in the CM input history database.

  The viewer information recording device 3 records the life information of the viewer monitor input based on the questionnaire from the input device 3a such as a keyboard or an optical reading device in the life time survey database 3b, and stores the preference information for the CM. Record in sensitivity database 3c.

  The questionnaire on daily life information and likability information distributes questionnaires to a large number of sample viewers selected on average from those in the area who can receive the same broadcast that meet certain conditions. It is done by a method that asks you to enter the actual state of viewing. Daily life information survey is a diary about whether or not you watched TV in each life time frame divided into 15 minutes in the broadcast time zone from 5:00 to 28:00 for each day on weekdays and holidays Ask for answers. Since the question does not specify a program or channel, it can be easily answered based on past memories.

  The life time survey database 3b includes monitor code and attribute information such as age, gender, and category (information indicating which category belongs to age and gender), and from 5:00 to 28:00 for each weekday / holiday. Information on whether or not the monitor was watching TV (viewing information), whether or not NHK was watched, for each life time frame (time frame divided into degrees) divided into 15 minutes Information (NHK viewing information), information on whether or not you were watching private broadcasting (public broadcasting viewing information), information whether you were dedicated to watching TV or other actions at the same time (viewing attitude information), and viewer monitor The fact (at-home information) or the like of whether or not he / she was at home in the life time frame is recorded. Here, NHK viewing information, commercial viewing information, viewing attitude information, at-home information, etc. are not used for calculating the viewer encounter level, but they are combined with other information to analyze the contact rate with consumers from various perspectives. Used for.

  In the Favorability Survey, about 5 of the CMs viewed in a certain period have a favorable impression, and the advertiser name, product name, features, etc. are answered in a descriptive manner. In addition, he / she is asked to enumerate favorable factors from the choices and answer in the form of YES / NO whether or not there is a trial preparation.

  The CM likability database 3c includes monitor code and attribute information such as age, gender, category, etc., the advertiser code of the CM work selected by the monitor, the target product code, and the work code, as well as each of the favorable factors presented in the survey form. Information on whether or not the customer has selected (favorable factor information), information on whether or not the product is ready for preparation (trial preparation information), information on whether or not the product is being used habitually (favorite information), Information indicating whether or not there is interruption (favorite interruption information) is recorded.

  The CM effect evaluation device 4 is based on the CM input history database 1e, the CM work database 2b, the life time survey database 3b, and the CM preference database 3c created above. With respect to the advertiser, the effect of the CM input in a certain period is analyzed by various methods, and the result is recorded in the CM evaluation database 4b in a format that can be called at any time or output from the output means 4c.

  The CM effect evaluation means 4a broadcasts CMs for each life time frame for various periods such as half-month, January, quarter, half-year, and year from the CM input history database 1e for each CM recorded in the CM work database 2b. Extract the history, calculate the broadcast frequency, extract viewing information from the life time survey database for each life time frame, and count the number of data with positive values to watch TV in that life time frame The number of viewer monitors is calculated to calculate the number of broadcasts, and the product of the number of broadcasts and the number of viewers is totaled for all life time frames to calculate the average number of viewer encounters (meet rate) of CM works. (Figure 3)

  The CM effect evaluation means 4a further extracts a record of the CM work selection by the viewer monitor from the CM likability database, a value obtained by dividing the number of selections by the total number of viewer monitors (favorite), Number of viewer monitors with positive values (favorability by factor), number of viewer monitors with positive test preparation information (number of trial intention votes), and viewing with negative test preparation information Number of viewer monitors (number of votes without trial intention), number of viewer monitors with positive value of favorite information (number of favorite votes), number of viewer monitors with positive value of favorite interrupt information (number of favorite interrupt votes) ) Respectively. In addition, the number of viewer monitors that have selected the CM work is counted for each category of viewer monitor (number of votes by level). Then, in association with the work code, advertiser code, target product code, and survey period, the work attribute information, the calculated broadcast frequency, the audience encounter level, the favorable feeling, the factor-specific favorable feeling, the number of trial preparation votes, The number of votes without trial intention, the number of favorite votes, the number of favorite suspended votes, and the number of votes by hierarchy are recorded in the CM evaluation database 4b. In addition to these, even if the work age acquired from the CM work database 2b, the GRP (Gross Rating Point) that is an index of the CM input amount calculated by a normal method from the program audience rating database (not shown), etc. are further recorded. good. The output unit 4c is an image display device, for example, and outputs various CM analysis results in the form of tables and graphs based on instructions from an auxiliary input unit (not shown).

  Of the CM evaluation database 4b created as described above, data relating to a specific business type or a specific advertiser is stored in a removable recording medium 5 such as a CD-ROM for analysis by an advertising expert or advertiser. Provided.

  Next, an example of an evaluation method using the broadcast CM effect evaluation apparatus will be described with reference to an actual output format. FIG. 4 shows a tabular report 7 output on the screen by the output means 4c. Each row of the table corresponds to one CM work and each column corresponds to an evaluation item of the CM work. All the main evaluation items are displayed for the combination of CM works selected by the operation from the input means (not shown). Yes.

  The display items in each column of the table will be described. The item 70 indicates which survey period the displayed data corresponds to, and the display is performed by an operation using a supplementary input device such as a mouse. Can be switched. Items 71 to 76 display information extracted from the CM evaluation database 4b for easy comparison. Reference numeral 71 denotes CM work information indicating information on the number of works aired on the advertiser, target product, month, and the same product. Reference numeral 72 represents the frequency of CM insertion, and indicates the number of broadcasts of each work and the rank among all CM works. Reference numeral 73 represents the degree of encounter with the viewer, and shows the average encounter degree (number of meet) of the viewer monitor of each work and the rank among all CM works. 74 represents the memory rate by the viewer, and shows the likability of each work and the ranking among all CM works. 75 represents the degree of purchasing behavior, and indicates the number of intentions to use each work and its rank among all CM works. Reference numeral 76 represents the degree of final use, which indicates the number of favorite votes for each work and the rank among all CM works. Item 77 is the overall evaluation and ranking of works calculated from these indices, but the calculation method is omitted.

  By using the TV CM effect evaluation device of the present invention, the effect of each broadcasted CM can be expressed objectively even if the viewing mode of the viewer changes, so the advertiser creates a CM dropping plan. Help. Furthermore, the value of the broadcast medium itself can be improved by objectively verifying the effect of introducing the broadcast CM.

The figure which shows the whole apparatus structure of the effect measuring apparatus of television CM which concerns on this invention. The block diagram which shows the flow of a process of the same apparatus. The flowchart which shows the average viewer encounter frequency (meet rate) calculation procedure of a present Example. The figure which shows an example of the evaluation report by the apparatus of this invention.

Explanation of symbols

1 CM input history creation apparatus 1a Reference information database 1b Information input means 1c Feature extraction means 1d CM detection means 1e CM input history database

2 CM work information recording device 2a Input means 2b CM work database

3 viewer information recording device 3a input device 3b life time survey database 3c CM favorability database

4 CM effect evaluation device 4a CM effect evaluation means 4b CM evaluation database 4c Output means

5 recording media

Claims (4)

The first recording means receives the broadcast, extracts the feature information of the audio or video, compares it with the reference information stored in advance in the reference information storage means, and stores the CM broadcast history including the matched CM identification information and the broadcast date and time. In addition, the viewing information for each life time frame collected from the viewer monitor is recorded in the second recording means, and the CM broadcast history for each life time frame is extracted from the first recording means for any CM. Then, the broadcast frequency is calculated, the number of viewer monitors who have watched the TV in the same life time frame is extracted from the second recording means, and the viewer is calculated from the product of the broadcast frequency and the number of people in all life time frames. TV commercial effect measurement method for calculating the average CM encounter degree and storing it in the third recording means together with the CM identification information . Reference information storage means for storing feature information of CM audio or video as original feature information, information input means for inputting voice or video information from a broadcasting device, and feature extraction for extracting feature information from the input voice or video information Means, CM detecting means for detecting that a specific CM is broadcast by comparing the extracted feature information with the stored original feature information, and identification information of the detected CM together with broadcast time information A CM insertion history creating unit having CM insertion history recording means for recording;
CM work information creation unit having input means and CM work recording means for recording CM identification information, advertiser identification information, target product identification information, work attribute information input from the input means,
Based on the input means and the input questionnaire results, the viewer monitor identification information, attribute information, and viewing information that is the fact that the viewer monitor was watching the surveyed broadcast in each life time frame are recorded. Viewer information recording unit having life information recording means
For the CM recorded in the CM work recording means, the CM broadcast history is extracted for each life time frame from the CM input history recording means to calculate the broadcast frequency, and the viewing information is extracted from the life recording means for each life time frame. The number of viewer monitors who were watching TV was calculated, and the product of airing frequency and number of people was summed over all time frames to calculate the CM viewer's encounter degree. CM effect evaluation unit having a CM effect evaluation unit that records the CM identification information, the advertiser identification information, and the target product identification information in a CM evaluation recording unit, and an output unit that outputs an evaluation result,
TV CM effect measuring device. Along with the inputted viewer monitor identification information and attribute information, any of the CM identification information of the CM selected by the viewer monitor as having a favorable feeling and a predetermined favorable factor for each selected CM Favorability recording means for recording likability factor information that is information as to whether or not there is a preparation tendency information that is information as to whether or not the preparation target of the CM target product is included,
The CM effect evaluation unit of the CM effect evaluation unit extracts a record of CM selection by the viewer monitor from the preference recording unit for each CM, and is a value obtained by dividing the selected number by the total number of viewer monitors. , The number of viewers with positive values for each likability factor information and the number of viewer intentions with the number of viewer monitors with positive values for trial preparation information. 3. The television CM effect measuring apparatus according to claim 2, wherein the sensitivities, the factor-specific sensitivities and the number of trial intention votes are recorded together with the CM identification information, the attribute information, and the audience encounter degree in the CM evaluation recording means.   Living information recording means of the viewer information recording unit, together with identification information of the viewer monitor, attribute information, and viewing information, home information that is the fact whether the viewer monitor was at home in each life time frame, and 4. The television CM effect measuring apparatus according to claim 2, further comprising: viewing attitude information that is a fact of whether or not each TV is dedicated to watching TV.

Images