JP2020188378A - Cm ordering automation system, program, and method - Google Patents

Abstract

To quickly execute determination of propriety of commercial message transmission order without depending on labor of a human while performing learning considering a balance of the whole CM broadcast frame of each time zone.SOLUTION: A CM ordering automation system acquires types and insertion conditions of CM broadcast frames; acquires types and insertion conditions of CMs; sets relevance factors of the respective CMs for each CM broadcast frame as broadcast frame penalty values and sets degrees of priority on order or interval of the CMs inserted into the CM broadcast frame and relevance factors based on the degrees of priority as order penalty values; calculates broadcast frame penalty values on the basis of the relevance factors of the respective CMs for each CM broadcast frame and settings by a penalty value setting unit 111c; and determines broadcast order of the CMs on the basis of a score value obtained by summing up the broadcast frame penalty values and the order penalty values each time.SELECTED DRAWING: Figure 1

Description

The present invention relates to a CM ranking automation system, program and method for automatically determining the suitability of the transmission order of broadcast commercials.

Conventionally, the ranking of CM insertion in TV broadcasting is assigned according to the contract information such as the broadcasting time zone provided by the advertising agency, and the broadcasting order (ranking) after the allocation is basically left to the broadcasting station. , It is done manually according to various rules and conditions. The suitability of this broadcast order is visually checked with a list showing the broadcast order.

Judgment as to whether or not such a CM broadcasting order is appropriate may be determined by the fact that there are times when a specific CM cannot be broadcast due to industry regulations, for example, alcohol-related commercials cannot be broadcast from 5:00 to 18:00 on weekdays. , At the time of CM contract, the end of the main part and the CM of the same sponsor are lined up, the position (time) of broadcasting etc. is specified by 0:00, and the alcohol related CM and the automobile related CM are not lined up, sponsors of the same industry It is necessary to take into account various conditions such as restrictions due to the mutual relationship between commercials such as not arranging.

By the way, as described above, conventionally, since the validity of the sending order is visually checked by a human being, not only a huge amount of time and labor is required, but also there is a problem that a check error may occur. doing. On the other hand, by checking the transmission order of commercials and correcting the transmission order when it is inappropriate by data processing using the AI method, it is possible to execute it quickly without human labor. A system has been proposed (see, for example, Patent Document 1). This system compares the attributes or program contents of consecutive commercials and programs, determines whether the transmission order is appropriate or inappropriate, and if it is determined to be inappropriate, the transmission order of the inappropriate part. Is rearranged, the determination is performed again, and the sorting is repeated until it is determined to be appropriate. In the technique described in Patent Document 1, the number of repetitions is set in advance so that the sorting process does not continue when an inappropriate part cannot be corrected.

Japanese Unexamined Patent Publication No. 7-21261

In the system disclosed in Patent Document 1 described above, the appropriateness / improperness of continuous commercials is determined and repeated a predetermined number of times until it becomes appropriate, and the portion that cannot be corrected even if the rearrangement is repeated a predetermined number of times. Stops sorting and outputs inappropriate parts to the display or printer.

However, in the system according to Patent Document 1 described above, only the suitability of continuous CMs is determined, and the uncorrectable parts are left as they are, and the inappropriate parts are partially pointed out. There is a problem that the degree of the inappropriate part cannot be judged from the comparison with other conditions and the overall balance. Further, although the system of Patent Document 1 has a mechanism for determining whether or not the transmission order is appropriate by the AI method using a knowledge database, it does not include information for evaluating the overall balance. , There was a certain limit to the evolutionary computation of AI.

Therefore, the present invention solves the above-mentioned problems, and an object of the present invention is to learn whether or not the commercial transmission order is appropriate while considering the balance of the entire CM broadcasting frame in each time zone. , To provide CM ranking automation systems, programs and methods that enable rapid execution without human effort.

In order to solve the above problems, the present invention automates CM ranking to set the broadcast order of CMs assigned to CM broadcast frames between programs when transmitting broadcast data related to the programs to a viewer's receiving device. In the system, a broadcast frame setting acquisition unit that acquires the type and insertion condition of a CM broadcast frame as broadcast frame setting information, a CM setting acquisition unit that acquires the type and insertion condition of a CM as CM setting information, and each CM broadcast. A broadcast frame penalty function that calculates the conformity rate of each CM with respect to the frame as a broadcast frame penalty value is set, and the order or interval priority of the CMs inserted in the CM broadcast frame and the conformance rate based on the priority are set as an order penalty. It is provided with a penalty value setting unit for setting an order penalty function to be calculated as a value.

Further, the present invention includes a CM insertion unit that inserts a CM into a CM broadcast frame and saves the insertion result, CM setting information of the CM inserted by the CM insertion unit into the CM broadcast frame, and the inserted CM broadcast frame. Based on the broadcast frame setting information, the broadcast frame penalty value is calculated using the broadcast frame penalty function, the CM insertion unit detects the order and interval of the CMs inserted into the CM broadcast frame, and the order penalty function is used. The processing by the penalty value calculation unit that calculates the order penalty value, the CM insertion unit, and the penalty value calculation unit is repeated multiple times, and the CM is broadcast based on the score value that is the sum of the broadcast frame penalty value and the order penalty value at each time. It is provided with a determination unit that determines the order.

Further, the present invention is a CM ranking automation method for setting a broadcast order of CMs assigned to CM broadcast frames between programs when transmitting broadcast data related to the programs to a viewer's receiving device. A setting acquisition step in which the broadcast frame setting acquisition unit acquires the broadcast frame type and insertion condition as broadcast frame setting information, and the CM setting acquisition unit acquires the CM type and insertion condition as CM setting information, and a penalty value setting unit. Set a broadcast frame penalty function that calculates the conformity rate of each CM for each CM broadcast frame as a broadcast frame penalty value, and is based on the priority of the order or interval of the CMs inserted in the CM broadcast frame and the priority thereof. It includes a penalty value setting step for setting an order penalty function for calculating the precision rate as an order penalty value.

Further, the CM ranking automation method of the present invention has a broadcast frame penalty function based on the CM setting information of the CM inserted into the CM broadcast frame by the CM insertion unit and the broadcast frame setting information of the inserted CM broadcast frame. The penalty value for calculating the broadcast frame penalty value using the above, the CM insertion unit detects the order and interval of the CM inserted in the CM broadcast frame, and the penalty value calculation unit calculates the order penalty value using the order penalty function. The calculation step and the processing by the CM insertion unit and the penalty value calculation unit are repeated a plurality of times, and the determination unit determines the CM broadcast order based on the score value which is the sum of the broadcast frame penalty value and the order penalty value at each time. With steps.

In the above invention, it is preferable that the penalty value setting unit sets the broadcast frame penalty value with a higher priority than the order penalty value. Further, in the above invention, the determination unit ends or completes the repeat process based on the number or duration of the repeat process by the CM insertion unit and the penalty value calculation unit, or the number or duration of the score values converged for each time. It is preferable to have a function of advancing to the next stage.

In the above invention, the CM insertion unit treats the insertion result at each time as one individual, associates the score values for each individual with each individual and accumulates them as a data set, and from a predetermined number of data sets, an individual with a predetermined score value. The next-generation individual by reallocating the individual extraction unit that extracts the data as an elite individual, some CM broadcasting frames included in the elite individual extracted by the individual extraction unit, and the CM inserted in the CM broadcasting frame. The next-generation generation unit is provided, and the determination unit uses a predetermined number of data sets as one population and sets the next population including the elite individuals and next-generation individuals extracted from the preceding population. From the data sets generated and included in the final population, it is preferable to determine the CM broadcasting order based on the score value of each data set.

In the above invention, the next-generation generation unit selects a plurality of arbitrary individuals from the preceding population, and combines some CM broadcasting frames included in the selected individuals and CMs inserted in the CM broadcasting frames. It is preferable to have a crossing function to generate a mutant individual.

The CM ranking automation system and the CM ranking automation method according to the present invention described above can be realized by executing the CM ranking automation program of the present invention described in a predetermined language on a computer. That is, the CM ranking automation program of the present invention is installed on a mobile terminal device, a smartphone, a wearable terminal, a mobile PC or other information processing terminal, an IC chip of a general-purpose computer such as a personal computer or a server computer, or a memory device, and is mounted on a CPU. By executing the above, it is possible to construct a CM ranking automation system having each of the above-mentioned functions and implement the CM ranking automation method according to the present invention.

Further, the CM ranking automation program of the present invention can be distributed through a communication line, for example, and is a package application that operates on a stand-alone computer by recording on a computer-readable recording medium. Can be transferred as. Specifically, the recording medium can be recorded on various recording media such as a magnetic recording medium such as a flexible disk or a cassette tape, an optical disk such as a CD-ROM or a DVD-ROM, or a RAM card. Then, according to the computer-readable recording medium on which this program is recorded, the above-mentioned system and method can be easily carried out by using a general-purpose computer or a dedicated computer, and the program can be stored, transported, and stored. Installation can be done easily.

As described above, according to these inventions, a score value is calculated based on the conformity rate of each CM with respect to each CM broadcast frame and the conformance rate of the order and interval of the CMs inserted in the CM broadcast frame, and the score thereof. It is possible to quickly determine the suitability of the commercial transmission order, which determines the CM broadcast order based on the value, without human effort, while learning in consideration of the balance of the entire CM broadcast frame in each time zone.

It is a conceptual diagram which shows typically the outline of the television broadcasting business in the television broadcasting station which concerns on embodiment. It is a block diagram which shows the internal structure of the information processing terminal which concerns on embodiment. It is a block diagram which shows the structure of the CM ranking automation system built on the information processing terminal which concerns on embodiment. It is a flow chart which shows the algorithm of AI processing in CM ranking which concerns on embodiment. It is explanatory drawing which shows the outline of the evolutionary computation in the AI processing which concerns on embodiment. It is explanatory drawing about evolutionary computation (order crossing) of AI processing which concerns on embodiment. It is explanatory drawing about evolutionary computation (mutation) of AI processing which concerns on embodiment. It is explanatory drawing which shows each value of the calculation parameter in the AI processing which concerns on embodiment.

Hereinafter, embodiments of the CM ranking automation system and the automation method according to the present invention will be described in detail with reference to the accompanying drawings. The embodiments shown below exemplify an apparatus or the like for embodying the technical idea of the present invention, and the technical idea of the present invention describes the material, shape, structure, and the like of each component. The arrangement etc. is not specified as the following. The technical idea of the present invention can be modified in various ways within the scope of claims.

(Flow of TV broadcasting business)
Explain the flow of work at TV stations and other institutions in TV broadcasting. FIG. 1 schematically shows an outline of the television broadcasting business in the present embodiment.
First, in response to an inquiry from an advertising agency, the sales office of a television station or the like submits a spot CM schedule table from the advertising agency side as shown in FIG. 1 according to a draft (S101). In this CM schedule table, the sales frame unit is divided for each program or time zone, and a cell which is a CM broadcasting frame is set for each sales frame unit. The advertising agency specifies the CM material to be assigned to each cell. According to this spot CM schedule table, the sales office of the TV station allocates the CM material to each cell (S102). After the allocation of the CM material, CM ranking is performed manually and by the AI engine (S103 and S104). When the CM ranking is made, the CM progress table is confirmed according to the CM ranking (S105), and when the CM material is brought in from the advertising agency (S106), the CM material is filed (S107) and goes on air. Preparations are underway for this.

On the other hand, the organization production station of the television station creates a program guide and gives a program production instruction to the production company (S112). In addition, the organization production station of the television station allocates the program material according to the created program guide (S109) and provides it for CM / program on-air (S109). The program material is brought in from the program production company (S113), and the program material is filed (S110). Then, after the program / CM progress table is confirmed (S105 and S109) and the CM material and the filing of the program material (S107 and S110) are completed, the CM / program is aired after the arrival of the broadcast time (S111). ..

(Composition of CM ranking automation system)
The CM ranking automation system according to the present invention is a system that sets the broadcast order of CMs assigned to CM broadcast frames between programs when transmitting broadcast data related to programs to a viewer's receiving device. In the embodiment, it is used in the CM ranking (S103 and S104) in the above-mentioned television broadcasting business. In the present embodiment, the CM ranking automation system and the automation method are realized by executing the CM ranking automation program of the present invention on the information processing terminal 1 as shown in FIG. That is, by installing the program of the present invention on a personal computer, a server device, a tablet PC or other general-purpose computer IC chip or a memory device as the information processing terminal 1 and executing the program on the CPU, each of the above-mentioned functions is provided. A system can be constructed to carry out the method according to the present invention.

Specifically, the information processing terminal 1 includes a storage 14 which is a storage unit, a CPU 11 which performs arithmetic processing, a memory 12 which temporarily stores information, and a communication interface 13 which communicates with the outside via the Internet 2. , Has an output interface 15 and an input interface 16. In the present embodiment, the storage 14, the CPU 11, the memory 12, the communication interface 13, and the like are connected to each other via the CPU bus 10, and data can be exchanged with each other.

The storage 14 is a device that stores data in a recording medium and reads out the stored data in response to a request from each device. For example, a hard disk drive (HDD), a solid state drive (SSD), a memory card, or the like is used. Can be configured.

The CPU 11 is a device that performs various arithmetic processes necessary for controlling each part, and by executing various programs such as an OS (Operating System), firmware, and an application program, various modules are virtually placed on the CPU 11. To construct. In the present embodiment, the CPU 11 executes a CM ranking automation program to make this general-purpose information processing terminal 1 function as a CM ranking automation system. The term "module" as used herein refers to a functional unit that is composed of hardware such as a device or device, software having that function, or a combination thereof, and for achieving a predetermined operation.

The communication interface 13 is a communication interface for performing data communication, and has a function of non-contact communication by wireless or the like, contact (wired) communication by a cable, an adapter means, or the like. It is a module that communicates with a wireless base station by wireless communication, and uses LTE lines, 3G / 4G lines, wireless communication networks such as Wifi (registered trademark) and Bluetooth (registered trademark), and other known devices and communication methods. be able to.

The CPU bus 10 is connected to an input interface 16 that accepts user operations and an output interface 15 that controls outputs such as video, sound, and LED lighting. The output interface 15 is a module that sends out video signals and audio signals in order to output video and audio from an output device such as a display and a speaker, and can output video and audio related to application software executed by the CPU 11. .. The input interface 16 is a module that receives operation signals from operation devices such as a touch panel and buttons, and the received operation signals are transmitted to the CPU 11 to perform operations on the OS and each application.

Then, the OS is started and executed on the CPU 11, and the basic functions of the information processing terminal 1 are managed and controlled by the OS. In addition, various applications can be executed on this OS, and various functions are executed by executing the applications.

In the present embodiment, the general-purpose information processing terminal 1 is made to function as a CM ranking automation system by executing a CM ranking automation program on the CPU 11. Specifically, as shown in FIG. 3, an information acquisition unit 111, a determination unit 112, a CM insertion unit 113, a penalty value calculation unit 114, and an output data generation unit 115 are constructed on the CPU 11.

The information acquisition unit 111 is a module that reads data from the input interface 16 and the storage 14 and performs various settings, and includes a CM setting acquisition unit 111a, a broadcast frame setting acquisition unit 111b, a penalty value setting unit 111c, and a calculation parameter setting. It has a part 111d.

The CM setting acquisition unit 111a is a module that acquires a CM type and insertion conditions as CM setting information. The broadcast frame setting acquisition unit 111b is a module that acquires the type and insertion condition of the CM broadcast frame as broadcast frame setting information.

This CM setting information is recorded in the master data stored in the storage 14, for example, as shown in FIG. 4, and includes sponsor competition information, negative competition information, special sponsor information, time regulation master, and basic conditions. Penalty master etc. are included.

The penalty value setting unit 111c sets a broadcast frame penalty function that calculates the conformity rate of each CM with respect to each cell (Cell []) CM broadcast frame as a broadcast frame penalty value, and orders the CMs inserted in the CM broadcast frame. Alternatively, it is a module that sets an order penalty function that calculates the priority of the interval and the precision rate based on the priority as the order penalty value. In the present embodiment, as shown in Table 1, the priority of the broadcast frame penalty value is set to be higher than the priority of the order penalty value.

For example, in the present embodiment, as shown in Table 1, the priority of the order or interval of CM, and the degree of conformance thereof are No. 1 as the order penalty value. 1 to 8 are set. The priority of these order penalty values is set in three stages of "high", "medium", and "low", and the penalty value which is the precision rate is set as the order penalty value according to the priority. .. In addition, as a broadcast frame penalty value, which is a conformity rate with respect to a broadcast frame, No. 9 to No. 14 is set, and these Nos. 9 to No. It is essential to satisfy the conditions for the 14 items, and the penalty value when the conditions are not satisfied is set to "110", which is higher than the other order penalty values.

The calculation parameter setting unit 111d is a module for setting break conditions, and sets the number of calculation seconds, the number of calculations, the number of current calculation stagnation seconds, the current number of calculation stagnation, and the like for the population processing to be repeatedly executed.

The determination unit 112 repeats the processing by the CM insertion unit 113 and the penalty value calculation unit 114 a plurality of times, and determines the broadcast order of the CM based on the score value which is the sum of the broadcast frame penalty value and the order penalty value in each time. Is.

Further, in the present embodiment, the determination unit 112 repeats based on the number or duration of the repeat processing by the CM insertion unit 113 and the penalty value calculation unit 114, or the number or duration of the score values converged for each time. It has a function to end the process or proceed to the next stage.

Further, the determination unit 112 uses a predetermined number of data sets as one population to generate the next population including the elite individual extracted from the preceding population and the next-generation individual, and finally the final individual. It has a function of determining the CM broadcasting order based on the score value of each data set from the data sets included in the group.

The CM insertion unit 113 is a module that inserts a CM into a CM broadcast frame and saves the insertion result. The CM insertion unit 113 treats the insertion result of each time as one individual, associates the score values for each individual with each individual, accumulates them as a data set, and elites the individuals with a predetermined score value from a predetermined number of data sets. The function as an individual extraction unit to be extracted as an individual, and some CM broadcasting frames included in the elite individual extracted by the individual extraction unit and CMs inserted into the CM broadcasting frame are reassigned to the next generation. It has a function as a next-generation generator that generates the individual.

The determination unit 112 uses a predetermined number of data sets as one population to generate the next population including the elite individual and the next-generation individual extracted from the preceding population, and includes it in the final population. The CM broadcasting order is determined based on the score value of each data set from the data sets.

Further, the function of the CM insertion unit 113 as a next-generation generation unit is to select a plurality of arbitrary individuals from the preceding population, and some CM broadcast frames included in the selected individuals and the CM broadcast frame thereof. It has a crossing function to generate a mutant individual by combining CMs inserted in.

The penalty value calculation unit 114 has a broadcast frame penalty based on the CM setting information of the CM inserted into the CM broadcast frame (Cell []) by the CM insertion unit 113 and the broadcast frame setting information of the inserted CM broadcast frame. The broadcast frame penalty value is calculated using the function, the CM insertion unit 113 detects the order and interval of the CMs inserted in the CM broadcast frame (Cell []), and the order penalty value is calculated using the order penalty function. It is a module.

The output data generation unit 115 is a module that generates as output file data, outputs the file, and prints out. Specifically, the CM material is assigned to the spot CM schedule table, and the CM progress information D1 in which the CM is ranked is output as the CM progress table. This CM progress table is provided for the CM material filing (S107) and the CM / program on-air (S111) shown in FIG.

(Operation of CM ranking automation system)
By operating the CM ranking automation system described above, the CM ranking automation method of the present invention can be implemented. In the present embodiment, the CM ranking automation method mainly uses the CM material allocation (S102) to the CM ranking (S103), AI ranking (S104), and CM progress table determination (S104) in the flow of the television broadcasting business shown in FIG. Used in S105). FIG. 4 shows the operation related to AI ranking by the CM ranking automation system. The processing procedure described below is only an example, and each processing may be changed as much as possible. Further, with respect to the processing procedure described below, steps can be omitted, replaced and added as appropriate according to the embodiment.

As shown in the figure, first, the maximum number of calculated seconds is set and the evaluation value is specified (S201). Next, according to the allocation in step S102, the CM setting information regarding the CM material is input to the cell (Cell []) (S202), and the suitability is evaluated for each cell based on the information of the CM material allocated to the cell (S202). S203). This conformity assessment acquires master data as broadcast frame setting information from the storage 14, and executes it based on this master data. This master data includes sponsor competition information, negative competition information, special sponsor information, time regulation master, and basic condition penalty master.

Specifically, in this conformity assessment, first, in the CM material allocation work in step S102, the broadcast frame setting acquisition unit 111b acquires the type and insertion condition of each cell input according to the spot CM schedule table as the broadcast frame setting information. Of this broadcast frame setting information, the cell type is an attribute or category assigned based on the program content, program genre, sponsor name, broadcast time zone, performer, target viewer group, etc. of each cell. Yes, it is input and set based on the spot CM schedule table produced by the advertising agency. Further, the insertion condition of each cell in the broadcast frame setting information is a CM insertion condition determined based on the cell type, for example, based on the time regulation zone included in the master data stored in the storage 14. The regulated CM types (for example, alcoholic beverages, games such as pachinko, consumer finance, etc.) are included. In addition, the CM insertion condition included in the master data includes a function to calculate the penalty value, and in addition to the type of each CM included in the cell, the position, order, appearance frequency, etc. in the cell are calculated. The formula and the coefficients in the formula are included. The function as the insertion condition of the CM is passed to the penalty value setting unit 111c as the insertion condition function.

Along with the acquisition of the broadcast frame setting information, the CM type and insertion condition assigned in step S202 are acquired as CM setting information. Among the CM setting information, the type of CM includes the type of business of each CM, the code (product), the sponsor name, and the number of playback seconds. Further, the insertion condition among the CM setting information is information set by reversely setting the above-mentioned broadcast frame setting information, and is a CM insertion condition determined based on the CM type. For example, the CM type (for example, , Alcoholic beverages, games such as pachinko, consumer finance, etc.) include time-controlled zones. The insertion condition of the CM setting information and the insertion condition of the broadcast frame setting information may be duplicated or distributed to both, and all the information is put together in one of them and the other is inserted. May be omitted.

Then, in the penalty value setting unit 111c, a function for calculating various penalty values is set. Specifically, the penalty value setting unit 111c sets a broadcast frame penalty function that calculates the conformity rate of each CM for each CM broadcast frame as a broadcast frame penalty value, and also sets the order of CMs inserted in the CM broadcast frame or Set an order penalty function that calculates the priority of the interval and the precision rate based on that priority as the order penalty value. The function set by the penalty value setting unit 111c is passed to the penalty value calculation unit 114.

Then, the population processing (S204) is executed by the AI engine. In this population processing, the result of the conformity assessment, the value set in step S201, the number of individuals and the number of threads are input. In the population processing (S204), first, the CM insertion unit 113 randomly inserts a CM into a cell which is a CM broadcasting frame, and executes a CM insertion step of saving the insertion result in the memory 12.

Next, the broadcast frame penalty value is calculated using the broadcast frame penalty function based on the CM setting information of the CM inserted into the CM broadcast frame by the CM insertion unit 113 and the broadcast frame setting information of the inserted CM broadcast frame. At the same time, the CM insertion unit 113 detects the order and interval of the CMs inserted in the CM broadcast frame, and executes a penalty value calculation step in which the penalty value calculation unit 114 calculates the order penalty value using the order penalty function.

After that, the processing by the CM insertion step and the penalty value calculation step is repeated a plurality of times until the break condition is reached (“N” in S205), and when the break condition is reached (“Y” in S205), the repeat processing is repeated. It proceeds to the end or the next stage, and executes a determination step of determining the broadcast order of the CM based on the score value which is the sum of the broadcast frame penalty value and the order penalty value at each time. In this determination step, the cell (Cell []) that is the execution result is written out, and the number or duration of the repeated processing by the CM insertion step and the penalty value calculation step, or the score value for each time is converged and the calculation is stagnant. The number of times or the duration is recorded as the break execution history (S206).

After the ranking by AI is completed, manual adjustment is performed as a final manual confirmation (S207). In this manual adjustment, the cell of the AI evolution calculation result (here, "Cell [] -α") is manually modified, and the modified result is saved as the modified cell (Cell [] -β). .. In this manual adjustment, the CM order is reviewed mainly based on the genre, performers, and the like. Then, after the final CM transmission order is determined by manual manual adjustment and the broadcast is actually performed, the AI learning engine performs the learning process based on the final broadcast result (S208). In this learning process, the cell "Cell [] -α" before modification is compared with the cell after modification (Cell [] -β) that has been manually modified and actually broadcast, and the changed condition is used as a change candidate. The accumulated conditions are created as a proposed change of the basic condition penalty master stored in the storage 14 (S209), and are reflected in the next AI ranking. As a result, the basic condition penalty master will be dynamically changed based on the broadcast performance.

Then, as shown in FIG. 3, CM progress information D1 is generated based on the CM ranking determined in this way, and the CM progress table is determined (step S105 in FIG. 1). This CM progress table is provided for CM material filing (S107) and CM / program on-air (S111). At this time, in the CM / program on-air (S111), the CM progress table, which is the result of this AI ranking, may be manually input on the commercial broadcasting system side for broadcasting the actual CM material, and is in a file format. It may be possible to take it into the commercial broadcasting system and automatically reflect it in the commercial broadcasting system.

(Evolutionary computation)
Here, the evolutionary computation in the population processing (S204) by the AI engine will be described in detail with reference to FIG. In this population processing, basically, a penalty value is specified for each condition, and "time to execute calculation" or "penalty value" is specified, and the order of CM is randomly rearranged to 100 types. Pattern (population) is created, and the one with the lowest score value, which is the sum of the penalty values, is selected. This is repeated a predetermined number of times, and the one with the lowest score value is determined as the final CM ranking. When generating these 100 types of patterns, some of the patterns with the lowest penalty are rearranged to create 100 new types, and the break condition "specified time" or "specified penalty value" is used. When the timing is reached, the optimum order at that time is selected and output.

More specifically, in the CM insertion step described above, the insertion result at each time is regarded as one individual, the score value for each individual is associated with each individual and stored in the memory 12 as a data set, and from a predetermined number of data sets, An individual extraction step of extracting an individual with a predetermined score value as an elite individual is executed, and some CM broadcasting frames included in the elite individuals extracted in the individual extraction step and CMs inserted in the CM broadcasting frame are re-examined. Perform the next generation generation step to perform the allocation and generate the next generation of individuals.

In this next-generation generation step in the present embodiment, a plurality of arbitrary individuals are selected from the preceding population, and as shown in FIG. 5, some CM broadcast frames included in the selected individuals and their CMs are selected. A mutation process is performed in which CMs inserted in the broadcast frame are combined to generate an ordered crossed individual or a mutant individual. The probability of causing this mutation processing is arbitrarily set as a constant.

Then, in the determination step, an elite individual and a next-generation individual (including a mutant individual) extracted from the preceding population are collected from a predetermined number of data sets as one population to form a next-generation population. Generate and calculate the score value of the dataset included in the next generation population. The generation of this next-generation population and the calculation of the score value are repeated until the break condition is reached.

Further, in the present embodiment, the break condition is the number of times or the duration of the calculation stagnant due to the convergence of the score values of the elite individuals as shown in FIG. 8 by monitoring the score values of the elite individuals in the population. In the example shown in the figure, the initial value of the score value at the start of calculation is set to 500 points and the evolution calculation is performed, and the score value of the elite individual decreases until the calculation number is 0.02 seconds and the number of calculations is the second. However, the number of calculation seconds is 0.03 seconds, and from the third calculation number, the decrease in the score value of the elite individual stops and the same value starts to be repeated, causing a stagnation of the calculation result. The calculation is terminated assuming that the break condition is reached when the stagnation of this calculation continues for a predetermined number of seconds or a predetermined number of times (in the illustrated example, the number of stagnation seconds is 10 seconds and the number of stagnations is 149 times).

After that, when the break condition is reached, the one with the lowest score value of each data set is picked up from the final population to determine the CM broadcasting order. In the example of FIG. 8, an individual having a score value of 300 points is determined as the final calculation result.

(1) Order crossing The above-mentioned order crossing will be described in detail. FIG. 6 shows columns p1 and p2 of the index of the spot frame included in the above-mentioned cell. As shown in FIG. 6, the indexes p1 and p2 have
p1 = (123/4567/89)
p2 = (452/1876/93)
It is included. Duplicate the central (between cutting points) element from each of these indexes.
c1 = (/ 4567 /)
c2 = (/ 1876 /)
Are generated respectively.

Next, the elements 934521876 ... X1 clockwise from the second cutting point (start position of 9) on p2.
Is extracted. At this time, if it protrudes from the right side, it is rotated to the beginning. Remove the same value as c1 from this X1
93218 ... X1'
And.

Further, using X1', the values are sequentially substituted from the second cutting point of c1, and c1 = (218/4567/93) which is a crossed individual.
To generate.

Similarly, an element 891234567 ... Y1 clockwise from the second cutting point of p1.
Is extracted, and the same value as c2 is removed from Y1.
92345 ... Y1'
Let Y1', and use this to sequentially substitute the values from the second cutting point of c2, and c2 = (345/1876/92), which is a crossed individual.
To generate.

(2) Mutation In mutation, as shown in FIG. 7, any one element is randomly replaced in the index columns p1 and p2 of the spot frame shown in FIG. 6, and a mutant individual c1 is generated. For example, in the same index, arbitrary columns 2 and 6 are exchanged and c1 = (164452789).
To generate.

(Action / effect)
As described above, according to the present embodiment, the broadcast frame penalty value is calculated based on the conformity rate of each CM for each CM broadcast frame, and the conformance rate of the order and interval of the CMs inserted in the CM broadcast frame is used. Calculate the order penalty value based on. Each of the penalty values is calculated by repeating the process a plurality of times under different conditions, and the broadcast order of the CM is determined based on the score value which is the sum of the broadcast frame penalty value and the order penalty value in each time. As a result, it is possible to quickly determine the suitability of the commercial transmission order without human effort while learning in consideration of the balance of the entire CM broadcast frame in each time zone.

In particular, in the present embodiment, since the broadcast frame penalty value is set to have a higher priority than the order penalty value, the ranking related to the broadcast frame penalty value in which time regulation such as CM of alcohols is indispensable. Is given priority, and the priority of ranking related to order penalties such as dispersion of competitors, which is relatively less important than this broadcast frame, is treated as low, and the balance of each time zone and the entire broadcast program is taken into consideration. Ranking is possible.

Further, in the present embodiment, the number or duration of repeated processing by the CM insertion unit 113 and the penalty value calculation unit 114, such as the number of calculation seconds / number of calculations and the number of stagnant seconds / number of times, or the score value for each time has converged. The number of times or the duration is set as a break condition, and when the break condition is reached, the operation repetition process ends or proceeds to the next stage, so that the operation process becomes too long or does not end due to an infinite loop. You can avoid the error.

Further, in the present embodiment, while generating the next generation by random and various algorithms, a plurality of individuals which are a data set of CM arrangement information and score value are generated, and the individuals in this population are sorted in ascending order of score value. However, since the top (for example, the top 10%) individuals are extracted as elite individuals and included in the next-generation population as they are, the next-generation with improved score values can be efficiently generated. Further, in the present embodiment, when a next-generation population is generated, a mutant individual that has undergone cross-processing between individuals is generated with a certain probability, so that it is difficult to generate from a combination of upper elite individuals. It is possible to create a new generation of new patterns and avoid meaningless stagnation of evolutionary computation.

D1 ... CM progress information 1 ... Information processing terminal 2 ... Internet 10 ... CPU bus 11 ... CPU
12 ... Memory 13 ... Communication interface 14 ... Storage 15 ... Output interface 16 ... Input interface 111 ... Information acquisition unit 111a ... CM setting acquisition unit 111b ... Broadcast frame setting acquisition unit 111c ... Penalty value setting unit 111d ... Calculation parameter setting unit 112 ... Determination unit 113 ... CM insertion unit 114 ... Penalty value calculation unit 115 ... Output data generation unit

Claims (15)

It is a CM ranking automation system that sets the broadcast order of CMs assigned to the CM broadcast frames between programs when transmitting broadcast data related to the programs to the viewer's receiving device.
A broadcast frame setting acquisition unit that acquires the type and insertion condition of the CM broadcast frame as broadcast frame setting information, and
A CM setting acquisition unit that acquires the CM type and insertion condition as CM setting information,
A broadcast frame penalty function that calculates the conformity rate of each CM for each CM broadcast frame as a broadcast frame penalty value is set, and the priority of the order or interval of the CMs inserted in the CM broadcast frame and the conformity rate based on the priority. Penalty value setting unit that sets the order penalty function, which calculates as the order penalty value,
A CM insertion unit that inserts a CM into the CM broadcast frame and saves the insertion result,
A broadcast frame penalty value is calculated using the broadcast frame penalty function based on the CM setting information of the CM inserted into the CM broadcast frame by the CM insertion unit and the broadcast frame setting information of the inserted CM broadcast frame. At the same time, the penalty value calculation unit that detects the order and interval of the CMs inserted into the CM broadcast frame by the CM insertion unit and calculates the order penalty value by using the order penalty function.
It is provided with a determination unit that repeats the processing by the CM insertion unit and the penalty value calculation unit a plurality of times and determines the broadcast order of the CM based on the score value that is the sum of the broadcast frame penalty value and the order penalty value at each time. CM ranking automation system featuring. The CM ranking automation system according to claim 1, wherein the penalty value setting unit sets the broadcast frame penalty value with a higher priority than the order penalty value. The determination unit ends or next repeats the process based on the number or duration of the repeat process by the CM insertion unit and the penalty value calculation unit, or the number or duration of convergence of the score value for each time. The CM ranking automation system according to claim 1, further comprising a function of advancing in stages. The CM insertion part is
The insertion result at each time is regarded as one individual, the score values for each individual are associated with each individual and accumulated as a data set, and the individual with a predetermined score value is extracted as an elite individual from a predetermined number of data sets. Individual extraction unit and
A part of the CM broadcasting frame included in the elite individual extracted by the individual extraction unit and the CM inserted in the CM broadcasting frame are reassigned to generate a next-generation individual. Prepare,
The determination unit uses the predetermined number of data sets as one population to generate the next population including the elite individual and the next-generation individual extracted from the preceding population, and finally the individual. The CM ranking automation system according to claim 1, wherein the broadcast order of CM is determined based on the score value of each data set from the data sets included in the group. The next-generation generation unit selects a plurality of arbitrary individuals from the preceding population, and suddenly combines some CM broadcasting frames included in the selected individuals and CMs inserted in the CM broadcasting frames. The CM ranking automation system according to claim 4, wherein the CM ranking automation system has a crossing function for generating a mutant individual. A CM ranking automation program that sets the broadcast order of CMs assigned to CM broadcast frames between programs when transmitting broadcast data related to the programs to the viewer's receiving device.
A broadcast frame setting acquisition unit that acquires the type and insertion condition of the CM broadcast frame as broadcast frame setting information, and
A CM setting acquisition unit that acquires the CM type and insertion condition as CM setting information,
A broadcast frame penalty function that calculates the conformity rate of each CM for each CM broadcast frame as a broadcast frame penalty value is set, and the priority of the order or interval of the CMs inserted in the CM broadcast frame and the conformity rate based on the priority. Penalty value setting unit that sets the order penalty function, which calculates as the order penalty value,
A CM insertion unit that inserts a CM into the CM broadcast frame and saves the insertion result,
A broadcast frame penalty value is calculated using the broadcast frame penalty function based on the CM setting information of the CM inserted into the CM broadcast frame by the CM insertion unit and the broadcast frame setting information of the inserted CM broadcast frame. At the same time, the penalty value calculation unit that detects the order and interval of the CMs inserted into the CM broadcast frame by the CM insertion unit and calculates the order penalty value by using the order penalty function.
The processing by the CM insertion unit and the penalty value calculation unit is repeated a plurality of times, and functions as a determination unit for determining the broadcast order of the CM based on the score value which is the sum of the broadcast frame penalty value and the order penalty value at each time. CM ranking automation program featuring. The CM ranking automation program according to claim 6, wherein the penalty value setting unit sets the broadcast frame penalty value with a higher priority than the order penalty value. The determination unit ends or next repeats the process based on the number or duration of the repeat process by the CM insertion unit and the penalty value calculation unit, or the number or duration of convergence of the score value for each time. The CM ranking automation program according to claim 6, further comprising a function of advancing in stages. The CM insertion part is
The insertion result at each time is regarded as one individual, the score values for each individual are associated with each individual and accumulated as a data set, and the individual with a predetermined score value is extracted as an elite individual from a predetermined number of data sets. Individual extraction unit and
A part of the CM broadcasting frame included in the elite individual extracted by the individual extraction unit and the CM inserted in the CM broadcasting frame are reassigned to generate a next-generation individual. Prepare,
The determination unit uses the predetermined number of data sets as one population to generate the next population including the elite individual and the next-generation individual extracted from the preceding population, and finally the individual. The CM ranking automation program according to claim 6, wherein the broadcast order of CMs is determined based on the score value of each data set from the data sets included in the group. The next-generation generation unit selects a plurality of arbitrary individuals from the preceding population, and suddenly combines some CM broadcasting frames included in the selected individuals and CMs inserted in the CM broadcasting frames. The CM ranking automation program according to claim 9, wherein the CM ranking automation program has a crossing function for generating a mutant individual. This is a CM ranking automation method that sets the broadcast order of CMs assigned to the CM broadcast frames between programs when transmitting broadcast data related to the programs to the viewer's receiving device.
A setting acquisition step in which the broadcast frame setting acquisition unit acquires the CM broadcast frame type and insertion condition as broadcast frame setting information, and the CM setting acquisition unit acquires the CM type and insertion condition as CM setting information.
The penalty value setting unit sets a broadcast frame penalty function that calculates the conformity rate of each CM for each CM broadcast frame as a broadcast frame penalty value, and the priority of the order or interval of the CMs inserted in the CM broadcast frame and the priority thereof. A penalty value setting step that sets an order penalty function that calculates the priority-based precision rate as an order penalty value, and
A CM insertion step in which the CM insertion unit inserts a CM into the CM broadcast frame and saves the insertion result,
The broadcast frame penalty value is calculated using the broadcast frame penalty function based on the CM setting information of the CM inserted into the CM broadcast frame by the CM insertion unit and the broadcast frame setting information of the inserted CM broadcast frame. At the same time, the penalty value calculation step in which the CM insertion unit detects the order and interval of the CMs inserted in the CM broadcast frame, and the penalty value calculation unit calculates the order penalty value using the order penalty function.
A determination step in which the processing by the CM insertion unit and the penalty value calculation unit is repeated a plurality of times, and the determination unit determines the broadcast order of the CM based on the score value which is the sum of the broadcast frame penalty value and the order penalty value in each time. CM ranking automation method characterized by being provided with. The CM ranking automation method according to claim 11, wherein in the penalty value setting step, the broadcast frame penalty value is set with a higher priority than the order penalty value. In the determination step, the repeat process is completed or next based on the number or duration of the repeat process by the CM insertion unit and the penalty value calculation unit, or the number or duration of convergence of the score value for each time. The CM ranking automation method according to claim 11, wherein the process is carried out in stages. The CM insertion step is
The insertion result at each time is regarded as one individual, the score values for each individual are associated with each individual and accumulated as a data set, and the individual with a predetermined score value is extracted as an elite individual from a predetermined number of data sets. Individual extraction step and
For some CM broadcast frames included in the elite individual extracted in the individual extraction step and CMs inserted in the CM broadcast frame, a next-generation generation step of executing reassignment to generate a next-generation individual is performed. Including
In the determination step, the predetermined number of data sets are used as one population to generate the next population including the elite individual and the next-generation individual extracted from the preceding population, and the final individual. The CM ranking automation method according to claim 11, wherein the broadcast order of CMs is determined based on the score value of each data set from the data sets included in the group. In the next-generation generation step, a plurality of arbitrary individuals are selected from the preceding population, and a part of the CM broadcasting frames included in the selected individuals and the CMs inserted in the CM broadcasting frames are combined and suddenly. The CM ranking automation method according to claim 14, further comprising a cross-generation step of generating a mutant individual.