JP7157267B1 - Advertisement Space Trading System, Advertisement Space Trading Method, and Computer Program - Google Patents

Abstract

An advertising space trading system for providing a platform for trading advertising spaces on a network improves the accuracy of recommending advertising spaces to users. A publishing unit publishes information on a plurality of advertising slots related to reserved advertisements on a network. The advertiser acquisition unit acquires advertiser information from the user. The purchase record acquisition unit acquires purchase record data corresponding to the advertiser information from each of the first purchase record database and the second purchase record database. The first purchase record database accumulates purchase record data of advertising spaces traded using the platform. The second purchase history database accumulates purchase history data of advertising spaces traded without using the platform. The recommendation unit recommends an advertising space to the user based on the purchase record data acquired from each of the first purchase record database and the second purchase record database. [Selection drawing] Fig. 6

Description

The present disclosure relates to an advertising space trading system, an advertising space trading method, and a computer program for providing a platform for trading advertising space on a network.

Patent Literature 1 discloses a system that publishes information on a plurality of advertising spaces across a plurality of advertising media on a network and accepts purchase requests for advertising spaces from users.

JP-A-2001-357294

In a platform for trading advertising space on a network, the types of advertising space traded are generally wide-ranging. Therefore, it is not easy for a user to select a desired advertising space on the platform.

Therefore, in order to assist the user in selecting an advertising space, it is conceivable to recommend the advertising space to the user, for example, using the purchase performance data of the advertising space traded using the platform. In this type of recommendation method, there may be a need to improve the accuracy of ad space recommendation to the user.

One aspect of the present disclosure aims to improve the accuracy of recommending advertising spaces to users in an advertising space trading system for providing a platform for trading advertising spaces on a network.

One aspect of the present disclosure is an advertising space trading system for providing a platform for trading advertising space on a network, comprising a publishing unit, an advertiser acquisition unit, a purchase history acquisition unit, and a recommendation unit. And prepare. The publishing unit is configured to publish information on a plurality of advertising slots related to reserved advertisements on the network. The advertiser acquisition unit is configured to acquire advertiser information, which is information about the advertiser, from the user. The purchase record acquisition unit is configured to acquire purchase record data corresponding to the advertiser information from each of the first purchase record database and the second purchase record database. The first purchase record database accumulates purchase record data of advertising spaces traded using the platform. The second purchase history database accumulates purchase history data of advertising spaces traded without using the platform. The recommendation unit is configured to recommend the advertising space to the user based on the purchase record data acquired from each of the first purchase record database and the second purchase record database.

According to such a configuration, it is possible to recommend an advertising space to a user by using not only purchase performance data of advertising space traded using the platform, but also purchase performance data of advertising space traded without using the platform. Therefore, it is possible to improve the accuracy of advertising space recommendation compared to a configuration in which an advertising space is recommended using only the purchase performance data of the advertising space traded using the platform.

It should be noted that the above users are assumed to be advertising space users such as advertisers and advertising companies that act as agents for advertisers' advertising activities.
In one aspect of the present disclosure, the advertising space trading system may further include a campaign acquisition unit. The campaign acquisition unit is configured to acquire campaign information, which is information about an advertising campaign, from a user. The recommendation unit may recommend the advertising space to the user based on the campaign information.

According to such a configuration, it is possible to recommend an advertising space suitable for the campaign information set by the user, thereby improving convenience for the user.
In one aspect of the present disclosure, the advertiser information may include information on the advertiser's industry or product or service category. Then, the purchase history acquisition unit obtains from each of the first purchase history database and the second purchase history database, the business type or category of the advertiser who purchased the advertising space specified by the advertiser information. You may acquire the purchase record data of the same advertising space.

According to such a configuration, it is possible to recommend an advertising space to the user based on the purchase performance data matching the advertiser's business type and category obtained from the user. As a result, it is possible to recommend an advertising space that better matches the user's needs.

In one aspect of the present disclosure, the second purchase record database accumulates the purchase record data of the advertising space identified from the advertisements that have already been placed as the purchase record data of the advertising space traded without using the platform. may be

According to such a configuration, due to the characteristic of advertising that the purchase record of the advertising space can be specified from the posted advertisement, the purchase record data of the advertising space traded without using the platform is collected and stored in the database. can. By using the purchase record data accumulated in this way, the ad space recommendation accuracy can be improved compared to the configuration in which the ad space is recommended using only the purchase record data of the ad space traded using the platform. can be improved.

In one aspect of the present disclosure, the advertising space trading system may further include a purchase candidate acquisition unit. The purchase candidate acquisition unit is configured to acquire information on a purchase candidate frame, which is an advertisement frame of a purchase candidate selected by the user. Then, the recommendation unit may recommend the advertising space to the user based on the information on the candidate purchase space.

According to such a configuration, it is possible to effectively utilize the information on the candidate purchase slots and recommend an appropriate advertisement slot to the user.
In one aspect of the present disclosure, the recommendation unit may preferentially recommend to the user an advertising space that is more likely to be purchased together with a purchase candidate space in the same advertising campaign.

According to such a configuration, it is possible to recommend to the user an advertising space that has a high probability of being purchased together with a candidate purchase space in the same advertising campaign, based on purchase record data of various advertisers.
In one aspect of the present disclosure, the recommendation unit may recommend to the user an advertising space that has the same or similar attributes as the purchase candidate frame.

According to such a configuration, it is possible to recommend to the user even an advertising space for which purchase record data is not yet available, such as an advertising space for a new medium.
In one aspect of the present disclosure, the recommendation unit uses the advertising space that the specific advertiser has purchased in the past in the advertising campaign of another advertiser who has the same business type or product or service category as the specific advertiser. Alternatively, an advertising space that is highly likely to be purchased together with an advertising space selected as a purchase candidate for an advertising campaign of a specific advertiser may be recommended with higher priority.

According to such a configuration, it is possible to recommend to the user an advertising space purchased in the same campaign by another advertiser who has the same business type or product service category as the specific advertiser.
In one aspect of the present disclosure, the advertising space trading system may further include a weighting unit. The weighting unit is configured to weight the first purchase quantity and the second purchase quantity. The first purchase amount is the purchase amount for one or more advertising slots of one or more advertisers, or an amount based on the purchase amount, specified based on the purchase record data acquired from the first purchase record database. is. The second purchase amount is the purchase amount for one or more advertising slots of one or more advertisers, or an amount based on the purchase amount, specified based on the purchase record data acquired from the second purchase record database. is. The recommendation unit may recommend an advertising space to the user based on the weighted first purchase amount and second purchase amount.

According to such a configuration, for example, when the first purchase amount is biased with respect to the second purchase amount, the bias can be corrected, and the ad space recommendation accuracy can be improved.
Another aspect of the present disclosure is an advertising space trading method for providing a platform for trading advertising space on a network, the method comprising: publishing information of a plurality of advertising spaces related to reserved advertising on the network. , obtaining advertiser information, which is information about the advertiser, from the user; a first purchase history database for accumulating purchase performance data of advertising spaces traded using the platform; and trading without using the platform. Acquisition of purchase record data corresponding to advertiser information from each of a second purchase record database that accumulates purchase record data of advertising spaces, and a first purchase record database and a second purchase record database and recommending advertising slots to users based on historical purchase data obtained from each of the.

Another aspect of the present disclosure is a computer program for causing a computer to function as an advertising space trading system for providing a platform for trading advertising space on a network, the publishing unit, the advertiser A computer is caused to function as an acquisition unit, the purchase record acquisition unit, and the recommendation unit.

According to these aspects, the same effects as those of the advertising space trading system described above can be obtained.

FIG. 1 is an explanatory diagram of the overall configuration of the advertising space trading system. FIG. 2 is a block diagram showing the configuration of the advertising space trading device. FIG. 3 is a diagram showing an example of the My Page screen. FIG. 4 is a diagram showing an example of a search screen. FIG. 5 is a diagram showing an example of an advertisement space information output screen. FIG. 6 is a flowchart of recommendation processing. FIG. 7 is a diagram showing an example of a recommendation condition setting screen. FIG. 8A is a schematic conceptual diagram of integrated data obtained by integrating the purchase record data acquired from the first purchase record database and the purchase record data acquired from the second purchase record database, and FIG. 8B is the integrated data. FIG. 4 is an explanatory diagram of a matrix generated from; FIG. 9 is a flow chart of the recommendation order determination process. FIG. 10 is a diagram showing an example of a recommendation result output screen.

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.
[1. Constitution]
The advertising space trading system 100 shown in FIG. 1 is a system for providing a platform for conducting advertising space trading on a network N such as the Internet (hereinafter referred to as "advertising space trading platform").

An advertising space transaction is a transaction between an advertising space provider who provides an advertising space and an advertising space user who uses the advertising space. Transactions between advertising space providers and advertising space users include not only direct transactions between advertising space providers and advertising space users, but also Transactions that are conducted indirectly through a third party with the user of the quota are also included.

Here, the advertising space is a transaction object, and refers to a space for inserting advertisements in media such as newspapers, magazines, radios, televisions, outdoor media, and digital (in other words, the Internet). Advertisement space users can post (for example, post, broadcast, distribute, etc.) advertisements in the ad space obtained through transactions. Outdoor media (hereinafter referred to as “ODM”) is a general term for media that come into contact with places other than home, such as transportation advertising media that come into contact with people in stations, trains, buses, etc. It includes outdoor advertising media that come into contact with the outdoors such as facilities.

In this embodiment, the advertising space provider is a media company or an advertising company acting on behalf of the media company's advertising space sales activities. Also, the advertising space user is an advertiser or an advertising company acting on behalf of the advertiser's advertising activities.

The advertising space trading system 100 is a system for trading advertising spaces for reserved advertisements among the advertising spaces. A reserved advertisement is an advertisement in which a specific advertisement space is reserved in advance, purchased, and placed. The reserved advertisement may be, for example, an advertisement in which the amount of money to be placed, the place to place the advertisement, the period of time to place the advertisement (for example, the amount of money to be placed, the place to be placed, the period of time to be placed), etc., are determined in advance.

Reserved advertisements include, for example, television commercials, radio commercials, newspaper advertisements, magazine advertisements, four-square advertisements, ODM advertisements (hereinafter referred to as "ODM advertisements"), digital advertisements, and the like. ODM advertisements include transportation advertisements and outdoor advertisements.

In the advertising space trading system 100, trading of advertising spaces of a plurality of media is performed in this way. In other words, transactions are made for a plurality of advertising spaces with different media.
The advertising space trading system 100 includes an advertising space trading device 1, a provider-side terminal 2, a user-side terminal 3, an advertising space information DB 4, a first purchase history database 5, and a second purchase history database 6. And prepare. Hereinafter, the database will be referred to as "DB". For convenience of illustration, one provider terminal 2 and one user terminal 3 are shown in FIG. , there are a plurality of each.
<Ad space transaction device>
The advertising space trading device 1 is a device that performs a central function of the advertising space trading system 100 and is a device for providing an advertising space trading platform. The advertising space trading device 1 is composed of a computer, specifically a server computer, and includes a communication section 11, a storage section 12, and a control section 13, as shown in FIG.

The communication unit 11 is a communication interface. The advertising space trading device 1 is connected to a network N such as the Internet as shown in FIG. I do. Also, the advertising space trading device 1 transmits and receives data to and from various DBs 4 to 6 via the communication unit 11 .

The storage unit 12 is storage such as a hard disk drive or a solid state drive, and stores various data.
The control unit 13 is mainly composed of a well-known microcomputer having a processor 13a such as CPU, semiconductor memory such as RAM, ROM, and flash memory (hereinafter referred to as "memory 13b"). Various functions of the control unit 13 are realized by the processor 13a executing a program stored in a non-transitional material storage medium. In this example, the memory 13b corresponds to a non-transitional substantive storage medium storing programs. Also, by executing this program, a method corresponding to the program is executed. Note that the number of microcomputers constituting the control unit 13 may be one or more.

In this embodiment, a program for generating the functions of the advertising space trading device 1 (hereinafter referred to as an advertising space trading program) is pre-installed in the storage unit 12 . By reading and executing this advertising space trading program by the processor 13a, the advertising space trading device 1 functions as a device that provides an advertising space trading platform. In addition, when the advertising space trading program is read and executed by the processor 13a, various data processing related to advertising space trading including the recommendation processing shown in FIG. 6, which will be described later, is executed.
<Provider side terminal>
The provider-side terminal 2 is a terminal owned by an advertising space provider, and is configured by a communication device such as a PC, tablet, or smart phone. The advertising space provider transmits information about the advertising space to be provided to the advertising space transaction device 1 through the provider-side terminal 2 . As a result, when the advertising space provider provides a new advertising space as an item to be traded, the advertising space provider informs the advertising space user, etc. of the information on the advertising space through the user-side terminal 3 and the advertising space trading device 1. is possible.
<User terminal>
The user-side terminal 3 is a terminal owned by an advertising space user, and is configured by a communication device such as a PC, tablet, or smart phone having a browsing function. The user-side terminal 3 includes a display and an input device that receives a user's input operation for information displayed on the display. The advertising space user receives information on the advertising space provided by each advertising space provider through the user-side terminal 3, and displays the received information on the display for confirmation. In addition, the user of the advertising space can select and apply for purchase of the advertising space at the user-side terminal 3 .
<Ad space information DB>
The advertising space information DB 4 stores information on advertising spaces traded on the advertising space trading platform (hereinafter referred to as “advertising space information”) and the like.

The advertising space information includes information such as a medium ID, a vehicle ID, a unit ID, an advertisement placement time, and a price of the advertising space.
The medium ID is an ID that identifies the type (classification) of advertising media such as newspaper, magazine, radio, television, ODM, and digital. A vehicle ID is an ID that identifies a vehicle. Here, the vehicle is the brand of the advertising medium. For example, the vehicle is the individual brand "XX Newspaper" for the advertising medium "Newspaper". A unit ID is an ID that identifies a unit. Here, a unit is a unit of advertising space for which a price is set in an advertising medium and sold by an advertising space provider. For example, in the advertising medium "newspaper", the unit is a sales unit such as "all 15 columns monochrome" or "2 columns 1/2 monochrome". The unit may specify at least one of the position, size and color of the advertisement on the advertising medium.

The advertisement placement time is the time when an advertisement is placed (published, broadcast, distributed, etc.) using an advertising space. In other words, the advertisement placement timing is the advertisement publication timing, the broadcasting timing, the distribution timing, and the like.
Note that the advertising space information may include information other than the above, and may include information such as an advertising space purchase application deadline and an advertisement submission deadline, for example.

The advertising space information is stored in the advertising space information DB 4 by each advertising space provider uploading data related to the advertising space via the provider-side terminal 2 . Then, the advertisement space information of each advertisement space stored in the DB 4 such as the advertisement space information is disclosed on the advertisement space trading platform as the information of the tradable advertisement space.

Information of each advertiser is stored in the advertisement space information DB 4 . Information on each advertiser is registered by an advertising space user or the like who uses the advertising space trading system 100 .
<First purchase record DB>
The first purchase record DB 5 accumulates purchase record data of advertising spaces traded using the advertising space trading platform. That is, when the advertising space is traded on the advertising space trading platform, the purchase history data of the traded advertising space is accumulated in the first purchase history DB 5 .

The purchase record data includes information on the purchased advertising space, information on the advertiser who purchased the advertising space, material information on the advertisement placed in the purchased advertising space, and the like.
The information on the purchased ad space includes information such as an ad space ID, which is information that enables identification of the ad space, the timing of placing the ad for the ad space, and the price of the ad space.

Here, the advertisement space ID may be, for example, a combination of the medium ID, the vehicle ID and the unit ID described above, or may be an ID corresponding to these combinations on a one-to-one basis. Also, the price of the advertising space here is the actual price of the advertising space when it is traded on the advertising space trading platform, unlike the estimated price described later.

The information on the advertiser who purchased the advertising space includes an advertiser ID that is information that can identify the advertiser. Further, the advertiser's information includes information (hereinafter referred to as "brand ID" etc.) that can identify the advertiser's brand, industry, product or service category, and the like. Here, examples of business types include automobile manufacturers and beverage manufacturers, and examples of product or service categories include automobiles and beverages. In addition, below, the category of goods or services is described as a "goods service category."

The advertisement material information includes, for example, information on the advertiser's product or service, information on the expression format of the advertisement, and the like. The advertisement expression format information includes, for example, type information such as text, image, and moving image, and color type information such as color and monochrome.
<Second Purchase Record DB>
The second purchase record DB 6 accumulates purchase record data of advertising spaces traded without using the advertising space trading platform. In other words, the second purchase record DB 6 accumulates the purchase record data of the advertising space traded between the advertising space provider and the advertising space user without going through the advertising space trading platform. The second purchase record DB 6 is constructed, for example, as follows.

In other words, due to the characteristics of advertising, it is possible to understand which advertiser used which advertising space, when and for what product or service, from advertisements that have already been placed (posted, broadcast, distributed, etc.). . In other words, it is possible to identify the purchase record of the advertising space from the published advertisement. In the second purchase record DB 6, purchase record data of advertising spaces specified from advertisements that have already been placed are accumulated.

In the present embodiment, the provider of the advertising space trading system 100 (hereinafter referred to as "system provider") acquires survey data including the purchase results of each advertising space specified from the advertisement placement, and uses the acquired survey data. Based on this, a second purchase record DB 6 is constructed. It is assumed that this survey data will be provided by an external research agency, which is a third-party organization. However, the provider of survey data is not particularly limited. For example, the survey data may be collected by the system provider itself, or part of the survey data may be provided by the ad space provider or the ad space user.

The purchase record data accumulated in the second purchase record DB 6 includes, in the same manner as the purchase record data accumulated in the first purchase record DB 5, information on the purchased advertising space, information on the advertiser who purchased the advertising space, It includes material information of the advertisement placed in the purchased advertisement space. However, in the second purchase record DB 6, an estimated price, which is a price estimated from the placement of advertisements, is stored as the price of the purchased advertisement space.

In other words, it is not possible to know how much the ad space was traded between the ad space provider and the ad space user from the published advertisement. Therefore, the survey data obtained by the system provider does not include the price of the traded ad space. Therefore, in the present embodiment, the system provider or the like calculates the estimated price of the advertising space from various information such as the type of the purchased advertising space, the type of business of the advertiser who purchased it, and the timing of placing the advertisement. 2 is stored in the purchase record DB 6. The method of calculating the estimated price is not limited to a method of human estimation based on various information, and may be a method of mechanically and automatically calculating based on various information.

In this embodiment, as will be described later, purchase record data for the same advertiser and advertising space is acquired from the first purchase record DB 5 and the second purchase record DB 6 . Therefore, name identification is performed for advertisers, advertisement spaces, and the like. Specifically, for the advertisement space ID, advertiser ID, brand ID, etc. stored in both DBs 5 and 6, common IDs that are commonly used in both DBs 5 and 6 are used.

Specifically, in this embodiment, when a new ad space or advertiser is registered on the ad space trading platform, the system provider collects survey data (that is, the second The same ID as the ID used in the purchase record DB 6) is assigned manually. Examples of such IDs include IDs that are commonly used in the advertising industry. In this way, the advertising space ID etc. (that is, common ID) used also in the second purchase history DB 6 are assigned to the advertising space ID etc. stored in the first purchase history DB 5 .

The second purchase record DB 6 and the first purchase record DB 5 store partially overlapping purchase record data. That is, as described above, the first purchase record DB 5 accumulates the purchase record data of the advertising space traded on the advertising space trading platform. Among them, purchase performance data of advertising spaces that have already been placed and purchase performance data of advertising spaces that have not yet been placed (in other words, advertising space purchased relatively recently) (Purchase performance data) and .

Of these, the former purchase record data is included in the aforementioned survey data after a certain amount of time has passed since the advertisement was placed. As a result, similar purchase record data is accumulated in the second purchase record DB 6 as well. In other words, the purchase record data of the ad space traded on the ad space trading platform for which a certain amount of time has passed since the placement of the ad is stored in both the first purchase record DB 5 and the second purchase record DB 6. accumulated. In this way, the second purchase record DB 6 may accumulate purchase record data of advertising spaces traded using the advertising space trading platform.

[2. function]
[2-1. Search/purchase application function]
Next, the advertising space search function and purchase application function, which are the basic functions of the advertising space trading system 100, will be described with reference to FIGS. 3 to 5. FIG. 3 to 5 are examples of screens displayed on the display of the user terminal 3. FIG. In the following description, it is assumed that an advertisement space user who owns the user-side terminal 3 is an advertising company that acts as an agent for advertisers' advertising activities.

First, the user, who is the person in charge of the advertising company, activates the user terminal 3 and accesses the website of the advertising space trading platform provided by the advertising space trading device 1 . Then, the user inputs authentication information such as a user ID and password to log in. Then, the My Page screen A1 shown in FIG. 3 is displayed on the display of the user terminal 3. FIG.

Various operation buttons are displayed on the My Page screen A1. Various operation buttons include a campaign creation button B1, a campaign search button B2, a purchase application button B3, and a search button B4. The campaign creation button B1 is a button for creating (registering) an advertiser's advertising campaign (hereinafter referred to as "campaign"). When the user operates the campaign creation button B1, a campaign creation screen (not shown) is displayed. A new campaign is registered by setting various information (for example, campaign name, advertiser name, advertiser's business type, advertised product or service category, campaign total budget and period, etc.) on the campaign creation screen. The registered campaign information is stored in the advertising space information DB 4, and the campaign name is displayed in the campaign display area R1 of the My Page screen A1. In this embodiment, the user selects a desired advertising space and adds the selected advertising space to the registered campaign. Then, by making a purchase application for the added advertising space, the purchase (that is, transaction) of the advertising space is performed.

The campaign search button B2 is a button for searching for a desired campaign from registered campaigns.
The purchase application button B3 is a button for making a purchase application for the advertising space added to the campaign. Here, in the advertising space display area R2 of the My Page screen A1, there are information about the advertising space added to the campaign (in this embodiment, information about the most recent advertising space added to the campaign) and a purchase application button B3. and is displayed. The displayed advertising space information includes, for example, a campaign name 21, a vehicle name 22, a unit name 23, an empty space status 24, a specific image 25 of the advertising space, a price 26, a purchase application deadline 27, and a submission deadline. 28 information. By operating the purchase application button B3, a purchase application for the advertising space displayed in the advertising space display area R2 is made.

On the other hand, the search button B4 is a button for searching for advertising space. When the user operates the search button B4, the screen transitions to the search screen A2 in FIG.
On the search screen A2, the user can select a search method for the advertising space. Media search, keyword search, and conditional search can be selected as a search method. The medium search is a search by medium, the keyword search is a search by keyword, and the conditional search is a search by conditions (for example, conditions such as target segment, total budget/period of campaign, area, etc.).

FIG. 4 shows an example of a screen when medium search is selected. In the medium search, when the user selects a medium from the medium selection area R3 on the left side of the screen, various vehicles corresponding to the selected medium are displayed in the vehicle display area R4. In the example shown in FIG. 4, "newspaper" is selected as the medium.

On the other hand, a media recommendation button B5 is displayed on the search screen A2. When the media recommendation button B5 is operated, recommendation processing is executed. The recommendation process is a process of recommending an advertising space to the user according to conditions set by the user. The recommendation function by this recommendation process is a characteristic function of the advertising space trading device 1, and will be described in detail later.

On the other hand, when the user selects any vehicle in the vehicle display area R4 of the search screen A2, the screen transitions to the advertising space information output screen A3 shown in FIG.
Note that in this embodiment, when a keyword search or conditional search is performed, one or more vehicles matching the conditions are displayed. Then, when the user selects a vehicle from among the displayed one or more vehicles, the screen transitions to the advertising space information output screen A3 shown in FIG.

The advertising space information output screen A3 displays information about the advertising space related to the vehicle selected on the search screen A2. Specifically, a vehicle name 31, a medium company name 32, a vehicle description 33, a specific image 34 of an advertisement space, an advertisement space list 35, and the like are displayed on the advertisement space information output screen A3.

The advertisement space list 35 displays a price and add campaign button B6 for each of one or more units. The user selects an advertising space from the advertising space list 35 and operates the add campaign button B6. Then, a list of ongoing campaigns is displayed, and the user selects the campaign to which the advertising space is to be added from the displayed list. Then, the ad space is added to the selected campaign.

When an ad space is added, information on the added ad space is displayed in the ad space display area R2 of the My Page screen A1 shown in FIG. Then, when the user operates the purchase application button B3, a notice of placing an order for the advertising space is transmitted to the advertising space provider, and processing for purchasing the advertising space is performed.

[2-2. Recommendation function]
Next, the advertising space recommendation function, which is a characteristic function of the advertising space trading system 100, will be described. The recommendation function is implemented by the controller 13 of the advertising space trading device 1 executing a recommendation process. This recommendation processing will be described below with reference to the flowchart of FIG. Note that the recommendation process is executed by operating the media recommendation button B5 on the search screen A2 shown in FIG.

First, the control unit 13 acquires recommendation condition information from the user-side terminal 3 in S101 (S represents a step). The recommendation condition information is information indicating conditions that are taken into account when the control unit 13 determines advertisement spaces to be recommended. In other words, the advertising space corresponding to the recommendation condition information is recommended to the user.

The control unit 13 acquires recommendation condition information as follows. That is, the control unit 13 transmits data for displaying the recommendation condition setting screen A4 shown in FIG. Then, a recommendation condition setting screen A4 is displayed on the display of the user-side terminal 3 that has received the data, and the user sets recommendation condition information on the setting screen A4.

In this embodiment, the recommendation condition information includes advertiser information and campaign information. The advertiser information is information about the advertiser, and includes, for example, the name of the advertiser, the type of business of the advertiser, product service category, and the like. The campaign information is information about the campaign, and includes, for example, the total budget of the campaign, the implementation period (start date and end date), and the like.

Here, the user who is the person in charge of the advertising company can set the advertiser information (for example, advertiser name, etc.) of any advertiser registered in the system 100 on the setting screen A4. That is, the user can set not only the advertiser information of the target advertiser but also the advertiser information of an advertiser different from the target advertiser on the setting screen A4. The term "target advertiser" as used herein refers to a specific advertiser among advertisers for whom an advertising space to be added to the advertiser's campaign is selected using the advertising space trading system 100 . Specifically, when the target advertiser uses the system 100 instead of the advertiser (such as a person in charge of an advertising company) as in the present embodiment, the advertiser directly uses the system 100 When using , it is the advertiser concerned.

As will be described later, an advertising space is recommended based on the purchase record data of the advertiser specified by the advertiser information input by the user. In other words, when the user decides which ad space to add to the target advertiser's campaign, the user can receive an ad space recommendation based on the purchase record data of an advertiser different from the target advertiser.

On the other hand, as the campaign information, basically, it is assumed that the campaign information of the target advertiser (for example, the advertiser in charge of the user who is the person in charge of the advertising company) is input.
When the recommendation condition information is set on the setting screen A4, data indicating the recommendation condition information is transmitted from the user-side terminal 3 to the advertising space trading device 1. FIG. By receiving the data, the control unit 13 of the advertising space trading device 1 acquires recommendation condition information.

In the following description, the advertiser identified by the advertiser name included in the recommendation condition information acquired in S101 will be referred to as a "set advertiser" and will be distinguished from the target advertiser.
Subsequently, in S102, the control unit 13 acquires purchase record data corresponding to the recommendation condition information from the first purchase record DB 5 and the second purchase record DB 6, respectively.

Here, the control unit 13 acquires purchase record data that matches the campaign information acquired in S101. Specifically, for example, the control unit 13 refers to the advertising space information DB 4 and acquires the advertisement placement time and price of each currently available advertising space. Then, the control unit 13 acquires from the DBs 5 and 6 the purchase record data of one or a plurality of advertising spaces whose advertisement placement time is included in the campaign period and whose price is equal to or less than the total budget.

In S102, the control unit 13 acquires purchase record data in which the business type or product service category of the advertiser who purchased the advertising space is the same as the business type or product service category included in the advertiser information acquired in S101. .

Further, the control unit 13 acquires the purchase record data of each advertising space from the DBs 5 and 6 so that the acquired purchase record data do not overlap. In the present embodiment, the control unit 13 acquires purchase record data by specifying the range of the placement time of the purchased advertising space and the range of the type of the advertising space (for example, the advertising space ID). Avoid duplication of performance data.

Specifically, the control unit 13 acquires the purchase record data included in the aforementioned survey data for which advertisements have already been placed, from the second purchase record DB 6 . In other words, the control unit 13 acquires the purchase record data of the advertising space for which the advertising period was in the past (excluding the most recent past).

On the other hand, from the first purchase record DB 5, the control unit 13 retrieves the purchase record data not yet stored in the second purchase record DB 6 among the purchase record data of the advertising space traded within the advertising space trading platform. get. In other words, the control unit 13 stores the purchase record data of the purchased advertisement space for which the advertisement has not yet been placed (that is, the purchase record data of the advertisement space whose advertisement time is in the future) and the Acquired from the first purchase record DB 5 is the purchase record data of the advertising space not yet included in the survey data (that is, the past purchase record data of the advertisement space having the most recent ad placement time).

In other words, the control unit 13 specifies the range of the placement time of the advertising space, thereby obtaining the purchase record data of the advertising space whose placement time is after the most recent past, that is, the advertising space whose placement time is after a specific time. It acquires at least from 1st purchase record DB5.

Here, in survey data, the types of media to be surveyed are limited, and not all types of media are covered. Therefore, there may be advertising spaces that are traded on the advertising space trading platform, but the survey data does not include purchase performance data.

Therefore, the control unit 13 determines that the purchase performance data of such advertising space is before the specific time point, that is, the purchase performance data that can be included in the survey data from the viewpoint of the advertisement presentation time. may also be acquired from the first purchase record DB 5 .

In this way, the control unit 13 may determine the range of purchase record data to be acquired from the DBs 5 and 6 by designating the range of the type of the advertising space in addition to the range of the advertising time of the advertising space.
The method of determining the purchase record data to be acquired from the DBs 5 and 6 is not limited to this, and the purchase record data may be acquired by designating only one of the range of the advertising period and the type of the advertising space. Also, the range of purchase record data to be acquired may be determined by other methods.

As described above, in this embodiment, the control unit 13 excludes the purchase performance data also stored in the second purchase performance DB 6 from the purchase performance data stored in the first purchase performance DB 5. to acquire purchase performance data. In other words, the control unit 13 acquires the purchase record data redundantly stored in both DBs 5 and 6 from the second purchase record DB 6 instead of from the first purchase record DB 5 .

Subsequently, in S103, the control unit 13 integrates the purchase record data acquired from the DBs 5 and 6 respectively. Specifically, as shown in FIG. 8A, the control unit 13 moves the purchase record data acquired from the first purchase record DB 5 and the purchase record data acquired from the second purchase record DB 6 in the row direction (that is, , the direction in which the purchase history data is arranged).

Here, as described above, the common ID is used for the advertiser ID and the advertising space ID included in the purchase record data acquired from both the DBs 5 and 6, and name identification has been completed. Data obtained by integrating the purchase record data of both DBs 5 and 6 is hereinafter referred to as "integrated data".

In FIG. 8A, the descriptions of "past", "future", and "recent past" mean that the placement times of advertising spaces related to purchase record data are past, future, and most recent past. .
Subsequently, in S104, the control unit 13 executes recommendation order determination processing. The recommendation order determination process is a process of determining the recommendation order of each advertising space for the user. An advertisement space with a higher recommendation order is preferentially recommended to the user. The recommendation order determination process will be described below with reference to the flowchart of FIG.

First, in S201, the control unit 13 performs processing such as data extraction and cross tabulation on the integrated data generated in S103, thereby forming a two-dimensional matrix (hereinafter referred to as a "purchase amount matrix") as shown in FIG. 8B. ) produces Y1.

The rows of the purchase amount matrix Y1 correspond to advertiser IDs, the columns correspond to advertisement space IDs, and the components correspond to the purchase amounts of advertisement spaces. That is, the (i, j) component of the purchase amount matrix Y1 is the purchase amount for the i-th advertiser's j-th advertising space. In the present embodiment, the number of times an advertiser has purchased an advertising space is exemplified as the purchase amount.

Subsequently, in S202, the control unit 13 performs matrix decomposition such as non-negative matrix factorization (NMF) or singular value decomposition (SVD) on the purchase amount matrix Y1 generated in S201, and performs low-rank approximation matrix (Hereinafter referred to as "low-rank approximation matrix") Y2 (not shown) is generated. This low-rank approximation matrix Y2 is generated as follows.

That is, it is assumed that the total number of advertisers is M, the total number of advertising spaces is N, and the purchase amount matrix Y1 generated in S201 is a matrix of M rows and N columns. In this purchase amount matrix Y1, "0" is set for all elements with no purchase record. Then, an appropriate integer D smaller than M and N is determined. Then, assuming a matrix P with M rows and D columns and a matrix Q with N rows and D columns, the control unit 13 calculates matrices P and Q that satisfy “Y˜P·Q′”. Here, "Q'" indicates the transposed matrix of matrix Q. Each row p of the matrix P obtained here can be regarded as a vector representing the characteristics of each advertiser, and each row q of the matrix Q can be regarded as a vector representing the characteristics of each advertising space. A matrix Y2 that satisfies Y2=P·Q′ is set as a low-rank approximation matrix Y2 of the purchase amount matrix Y1. In addition, in the low-rank approximation matrix Y2, even an element with no purchase record can have a non-zero value. Therefore, by recommending an advertising space using the low-rank approximation matrix Y2, it is possible to recommend an advertising space that has not been purchased.

Subsequently, in S203, the control unit 13 acquires the fitness of each advertising space based on the low-rank approximation matrix Y2 generated in S202. The suitability here is the degree of suitability of the ad space for the recommendation condition information. In this embodiment, the control unit 13 uses the component of each column (that is, each ad space) of the row corresponding to the set advertiser included in the recommendation condition information in the low-rank approximation matrix Y2 as the suitability of each ad space. get. The suitability acquired in S203 is used when determining the recommendation order of each advertising space in S205, which will be described later.

Subsequently, in S204, the control unit 13 acquires the purchase amount of each advertising space of the specific company according to the advertiser information acquired in S101 of FIG. Specifically, the control unit 13 acquires the purchase amount of each advertising space of a specific company according to the type of business and product/service category included in the advertiser information. In this embodiment, as the specific company corresponding to the type of industry or product/service category, a predetermined major company in the type of industry or product/service category is assumed.

That is, in the present embodiment, corresponding specific companies (in other words, corresponding major companies) are set in advance for each industry or product/service category. Then, the control unit 13 acquires the components of each column (that is, each advertising space) of the row corresponding to the specific company (that is, the predetermined major company) in the purchase amount matrix Y1 generated in S201. As a result, the purchase amount of each advertising space of the specific company is acquired.

In this embodiment, the purchase amount of each advertising space of a specific company is obtained from the purchase amount matrix Y1, but the method of obtaining the purchase amount of a specific company is not limited to this. For example, in the low-rank approximation matrix Y2, the component of each column (that is, each advertising space) in the row corresponding to the specific company may be obtained as the purchase amount of each advertising space of the specific company.

Subsequently, in S205, the control unit 13 determines the recommendation order of each ad space. Here, the control unit 13 uses two methods of advertising based on the suitability of each advertising space acquired in S203 and the purchase amount (i.e., purchase record) of each advertising space of the specific company acquired in S204. Determine the order of recommendation for frames.

Specifically, first, the control unit 13 determines the recommendation order of each ad space so that the recommendation order of the ad space is higher for the ad space with a higher degree of suitability for the set advertiser acquired in S203. As a result, the recommendation order based on the degree of conformity is determined.

In addition, the control unit 13 determines the recommendation ranking of each advertising space such that the higher the purchase amount of each advertising space of the specific company acquired in S204, the higher the recommendation ranking. As a result, the order of recommendation based on the purchase record of the specific company is determined.

After executing the process of S205, the control unit 13 ends the recommendation order determination process of FIG.
Returning to FIG. 6, in S105, the control unit 13 executes presentation processing. The presentation process is a process of presenting to the user at least one tradable ad space stored in the ad space information DB 4 . Specifically, the control unit 13 acquires the advertising space information of one or more advertising spaces having relatively high recommendation rankings from the advertising space information DB 4 in accordance with the recommendation ranking determined in S205 of FIG. Then, it generates data for sorting the obtained advertising space information according to the recommendation order and displaying the list, and transmits the data to the user-side terminal 3 . A recommendation result output screen A5 shown in FIG. 10 is displayed on the display of the user-side terminal 3 that has received the data.

The recommendation result output screen A5 displays a recommendation result based on the degree of suitability for the set advertiser and a recommendation result based on the purchase record of a specific company corresponding to the set business type or product service category. Specifically, a plurality of pieces of advertising space information sorted in descending order of suitability are displayed in the upper part of FIG. On the other hand, a plurality of pieces of advertising space information sorted in descending order of purchase record of specific companies corresponding to the type of industry or product/service category set by the user are displayed in the lower part of FIG. In this embodiment, it is assumed that the user has selected an automobile manufacturer or automobile as the advertiser's industry or product service category. Therefore, on the output screen A5, a recommendation result based on the purchase record of a major automobile company is displayed as a specific company.

It should be noted that in the present embodiment, the degree of conformity and the value of the purchase record of a specific company are not displayed on the output screen A5. That is, these values are used on the backend side of the system to determine the recommendation ranking of each ad slot.

On the output screen A5, the user selects one of the plurality of ad spaces 41 and adds it to the target advertiser's campaign. Then, the user operates the purchase application button B3 on the My Page screen A1 of FIG. 3 to purchase the corresponding advertising space.

After executing S105, the control unit 13 ends the recommendation process shown in FIG.
[3. effect]
According to the embodiment detailed above, the following effects are obtained.

(1) In this embodiment, the control unit 13 acquires purchase history data corresponding to advertiser information acquired from the user from each of the first purchase history DB 5 and the second purchase history DB 6 . The first purchase record DB 5 accumulates purchase record data of advertising spaces traded using the advertising space trading platform. The second purchase record DB 6 accumulates purchase record data of advertising spaces traded without using the advertising space trading platform. Then, the control unit 13 recommends an advertising space to the user based on the purchase record data obtained from the first purchase record DB 5 and the second purchase record DB 6 respectively.

Therefore, it is possible to recommend an advertising space to a user by using not only purchase performance data of advertising space traded using the advertising space trading platform but also purchase performance data of advertising space traded without using the advertising space trading platform. Therefore, it is possible to improve the accuracy of ad space recommendation compared to a configuration in which an ad space is recommended using only the purchase performance data of the ad space traded using the ad space trading platform.

(2) In the present embodiment, the user can set advertiser information for an advertiser different from the target advertiser on the setting screen A4 for setting advertiser information. Then, the control unit 13 acquires the advertiser information from the user terminal 3 . In other words, the control unit 13 selects a user from among a plurality of advertisers registered in the advertising space trading system 100 (specifically, a plurality of advertisers whose information such as advertising space information is stored in the DB 4). obtains advertiser information for advertisers arbitrarily selected by Then, the control unit 13 recommends an advertising space based on the acquired advertiser information.

Therefore, the user can receive a recommendation of the same advertising space from an advertiser different from the target advertiser. In particular, in general online shops where daily necessities are purchased, users do not know what kind of products other specific users have purchased. It is not possible to receive product recommendations based on data.

On the other hand, due to the characteristics of advertisements, purchase records of other advertisers are specified from published advertisements. In other words, the purchase history of other advertisers is disclosed. Under these circumstances, in the advertising space trading system 100, the user can set an advertiser different from the target advertiser and receive a recommendation of advertising space based on the purchase performance data of the set other advertiser. .

(3) The control unit 13 acquires the campaign information included in the recommendation condition information from the user in S101, and recommends the advertising space to the user based on the acquired campaign information. Therefore, it is possible to recommend an advertising space suitable for the campaign information set by the user, thereby improving the user's convenience.

(4) In S102, the control unit 13 selects from each of the first purchase record DB 5 and the second purchase record DB 6 whether the type of business or product service category of the advertiser who purchased the advertising space is included in the recommendation condition information. Acquire purchase record data that is the same as the industry or the category specified by the main information. Therefore, it is possible to recommend an advertising space to the user based on the purchase performance data that matches the advertiser's business type and category acquired from the user. As a result, it is possible to recommend an advertising space that better matches the user's needs.

(5) The second purchase record DB 6 accumulates the purchase record data of the advertising space identified from the advertisements that have already been placed as the purchase record data of the advertising space traded without using the advertising space trading platform. be.

Therefore, due to the characteristic of advertisement that the purchase record of the advertising space can be identified from the published advertisement, the purchase record data can be collected and accumulated in the database even for the advertising space traded without using the platform. By using the purchase record data accumulated in this way, compared to a configuration in which an ad space is recommended using only the purchase record data of the ad space traded using the ad space trading platform, the ad space can improve the recommendation accuracy of

In this embodiment, the advertising space trading platform corresponds to the platform, S101 corresponds to the processing of the advertiser acquisition unit and the campaign acquisition unit, S102 corresponds to the processing of the purchase history acquisition unit, and S104 and S105. corresponds to the processing of the recommendation unit, and S105 corresponds to the processing of the disclosure unit.

[4. Other embodiments]
Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments, and can be implemented in various modifications.

(1) The purchase amount matrix Y1 generated in S201 of FIG. 9 is not limited to that of the above embodiment. For example, based on the integrated data generated in S103 of FIG. 6, the average purchase amount of each advertising space is calculated for each advertiser's business type or product service category. Then, a matrix may be generated as the purchase amount matrix Y1, in which the rows correspond to the IDs of the business types or product service categories, the columns correspond to the advertising space IDs, and the components are the average purchase amounts. Then, each advertisement space may be recommended using the generated matrix Y1.

(2) In the above embodiment, the control unit 13 acquires information on a purchase candidate frame, which is an advertisement space for a purchase candidate selected by the user, and recommends an advertisement frame to the user based on the acquired information on the purchase candidate frame. You may The purchase candidate frame here may be an advertising frame that has been added to the target advertiser's campaign but has not yet been offered for purchase. It should be noted that the process of acquiring the information of the candidate purchase frame corresponds to the process of the candidate purchase acquisition unit. According to such a configuration, it is possible to effectively utilize the information on the candidate purchase slots and recommend an appropriate advertisement slot to the user.

Specifically, for example, when the user selects an ad space to be added to the target advertiser's campaign, the ad space may be recommended using the information on the candidate purchase space, as in Examples 1 to 3 below. .
<Example 1>
The control unit 13 may exclude an advertising space that satisfies either one or both of the following conditions 1 and 2 from among the advertising spaces to be recommended (for example, the advertising space for which the recommendation order has been determined).
(Condition 1) Already selected as a purchase candidate frame (Condition 2) The total amount of one or more advertisement frames selected as a purchase candidate frame is subtracted from the total campaign budget set by the user. The advertising space must be priced above the remaining budget <Example 2>
The control unit 13 may preferentially recommend to the user an advertising space with a higher probability of being purchased together with a purchase candidate frame in the same advertising campaign (hereinafter referred to as "combination purchase rate").

For example, the control unit 13 refers to the purchase record data of the DBs 5 and 6, and ranks combinations of candidate purchase frames and advertisement spaces purchased in the same campaign based on purchase amounts. Then, the control unit 13 may recommend to the user an advertising space related to a combination with a high ranking.

For example, for advertising space A, which is a candidate purchase space, the combinations of advertising spaces that have been purchased in the same campaign as advertising space A and the number of purchases are as follows.
Purchase advertising space A and B as a set 5 times Purchase advertising space A, B and C as a set 3 times Purchase advertising space A and D as a set 2 times In this case, the number of times advertising space A and B are purchased as a set is high, advertising space B may be recommended to the user.

According to such a configuration, it is possible to recommend to the user an advertising space with a high concurrent purchase rate (in other words, a synergistic effect) with a candidate purchase space in the same campaign, based on the purchase performance data of various advertisers.

Note that for multiple ad slots traded using the inventory trading platform, system 100 can determine whether they were purchased for the same campaign. On the other hand, for multiple ad spaces traded without using the ad space trading platform, the system 100 may not be able to determine whether they were purchased for the same campaign.

Therefore, for multiple ad spaces traded without using the ad space trading platform, for example, the advertiser identified from each advertisement placed using those ad spaces, the timing of advertisement placement, the product to be advertised, or A plurality of ad slots with the same service, brand, etc. may be presumed to be a plurality of ad slots purchased for the same campaign.
<Example 3>
The control unit 13 may recommend to the user an advertising space that has the same or similar attributes as the candidate purchase space. Specifically, a tag indicating the attribute of each ad space (hereinafter referred to as "attribute tag") is attached to each ad space, and an advertisement having an attribute tag that is the same as or similar to the attribute tag of the purchase candidate space Frames may be recommended to the user. It should be noted that the attribute of the advertisement space is assumed to be the type of medium, vehicle, unit, etc., the name of the medium company, and the like.

The attribute tag may be manually attached to each advertisement space by a system provider, an advertisement space provider, or the like, or may be mechanically attached. In the latter case, for example, the relationship between the ad space information of each ad space and the attribute tag is learned by machine learning, etc., and the attribute tag is automatically attached to each ad space using the learned machine learning model. good.

According to such a configuration, an advertising space having the same or similar attributes as the purchase candidate frame is recommended to the user. Therefore, it is possible to recommend to the user even an advertising space for which purchase record data is not yet available, such as an advertising space for a new medium.

(3) Also, for example, an advertising space may be recommended by the following method. Specifically, for example, the control unit 13 acquires information indicating a specific advertiser that is a specific advertiser. The specific advertiser here is assumed to be the target advertiser, but the specific advertiser may be a set advertiser or the like other than the target advertiser.

Then, based on the purchase record data accumulated in the first purchase record DB 5 and the second purchase record DB 6, the control unit 13 selects campaigns of other advertisers having the same industry or product service category as the specific advertiser. , priority is given to ad slots that have a high probability of being purchased together with ad slots that have been purchased by a specific advertiser in the past or that have been selected as purchase candidates for an advertising campaign of a specific advertiser (combination purchase rate). may

For example, assume that the business type of the specific advertiser is an automobile manufacturer, and the specific advertiser has purchased advertising space A in the past or has selected it as a candidate purchase space. In this case, advertising spaces B and C having a high concurrent purchase rate with advertising space A may be recommended in campaigns of other companies of the automobile manufacturer.

According to such a configuration, it is possible to recommend to the user an advertising space purchased in the same campaign by an advertiser who has the same business type or product service category as the specific advertiser.
(4) In S102 of FIG. 6 of the above embodiment, the control unit 13 specifies the business type or product service category of the advertiser who purchased the advertising space from each of the DBs 5 and 6 by the advertiser information set by the user. Purchasing performance data that is not identical to the industry or category described above may also be obtained. Then, the control unit 13 may recommend an advertising space using the acquired purchase record data.

(5) In S101 of FIG. 6 of the above embodiment, the campaign information acquired from the user may include information indicating the conditions regarding the medium of the advertising space. Specifically, for example, the campaign information may include information specifying a specific medium, information excluding a specific medium, and the like. Then, an advertising space that matches the information included in the campaign information may be recommended.

(6) In S102 of the above embodiment, the control unit 13 excludes the purchase performance data also stored in the first purchase performance DB 5 from the purchase performance data stored in the second purchase performance DB 6. purchase record data may be obtained by In other words, the control unit 13 may acquire the purchase record data redundantly stored in both the DBs 5 and 6 from the first purchase record DB 5 instead of the second purchase record DB 6 .

(7) In the above embodiment, the "purchase amount", which is a component of the purchase amount matrix Y1, is the number of times the advertiser has purchased the advertising space, but the purchase amount is not limited to this.
For example, the purchase volume is the number of campaigns in which the advertiser used the ad space, the amount of the ad placed by the advertiser using the ad space, and the amount of the ad placed by the advertiser using the ad space. may be the number of times of contact or the like. In addition, as the number of advertisement contact times, for example, GRP (Gross Rating Point), the number of advertisement exposure times, or the like may be used.

(8) In the above embodiments, common IDs for advertising spaces, advertisers, etc. are manually assigned by the system provider. However, the method of assigning the common code is not limited to this. For example, a common code may be assigned mechanically. Specifically, for example, text matching is performed on the advertising spaces and advertiser names stored in both DBs 5 and 6 . Then, the system may automatically assign a common ID to the matching ad spaces and advertisers.

(9) On the output screen A5 shown in FIG. 10, values such as the suitability of each advertising space and the purchase amount of a specific company may be displayed.
(10) In the above embodiment, when generating the purchase amount matrix Y1 based on the purchase performance data acquired from the first purchase performance DB 5 and the purchase performance data acquired from the second purchase performance DB 6, The purchase amount of each advertising space of each advertiser specified from the former purchase record data and the purchase amount of each advertisement space of each advertiser specified from the latter purchase record data are treated equally. However, how to handle both purchase amounts is not limited to this. For example, both purchase amounts may be weighted, and the purchase amount matrix Y1 or the like may be generated based on the weighted purchase amounts.

In other words, the control unit 13 may weight the first purchase amount and the second purchase amount. Here, the first purchase amount is the purchase amount for one or more advertising spaces of one or more advertisers (in other words, each advertiser purchase amount for each advertising space), or an amount based on the purchase amount. The second purchase amount is the purchase amount for each of one or more advertising spaces of one or more advertisers specified from the purchase record data acquired from the second purchase record DB 6 (in other words, each advertiser's amount of purchase for advertising space), or an amount based on the purchase amount. The first purchase amount and the second purchase amount include the purchase amount of the advertiser specified by the advertiser information acquired from the user in S101 of FIG. 6 (that is, the set advertiser). Then, the control unit 13 may recommend an advertising space to the user based on the weighted first purchase amount and second purchase amount.

As an example, the control unit 13 may perform weighting as follows. That is, the control unit 13 calculates a weighted purchase amount matrix according to Equation 1 below.

Weighted Purchase Quantity Matrix=First Purchase Quantity Matrix+α×Second Purchase Quantity Matrix Equation 1

Here, the first purchase amount matrix (second purchase amount matrix) is obtained by performing data extraction, cross tabulation, etc. on the purchase performance data acquired from the first purchase performance DB 5 (second purchase performance DB 6). A matrix representing the purchase amount for each advertising space of each advertiser generated by performing processing. Similar to the purchase amount matrix Y1 described above, the rows of the first purchase amount matrix (second purchase amount matrix) correspond to the advertiser ID, the columns correspond to the advertisement space ID, and the components correspond to the purchase amount of the advertisement space. may The first and second purchase amount matrices correspond to the first and second purchase amounts, respectively. Also, α is a parameter.

Then, based on the weighted purchase amount matrix calculated according to Equation 1 above, the control unit 13 may generate a low-rank approximation matrix in the same manner as in the above embodiment, and may recommend an advertising space to the user.
In addition, the control unit 13 predicts the advertising space to be purchased by the advertiser (in other words, the advertising space for which the purchase application is accepted from the user) using the weighted purchase amount matrix, and sets the parameters so that the prediction accuracy is maximized. α may be determined. In other words, the control unit 13 may determine the parameter α so as to maximize the probability that the advertiser will purchase the advertising space recommended using the weighted purchase amount matrix.

Here, the control unit 13 executes the task of predicting the advertising space purchased by the advertiser based on the past purchase record data accumulated in the DBs 5 and 6, and determines the parameter α so as to maximize the prediction accuracy. You may Alternatively, the control unit 13 actually recommends an advertising space to the user with various values of the parameter α, and determines the parameter α based on performance data such as whether or not the recommended advertising space has been purchased. may That is, the control unit 13 may determine the parameter α through actual operation.

According to such a configuration, for example, when the first purchase amount is biased with respect to the second purchase amount, the bias can be corrected, and the ad space recommendation accuracy can be improved.
Specifically, for example, the purchase performance data of the first purchase performance DB 5 and the purchase performance data of the second purchase performance DB 6 may differ in nature.

Specifically, the purchase record data of the second purchase record DB 6 based on the survey data has higher completeness than the purchase record data of the first purchase record DB 5 based on transactions on the advertising space trading platform, and more May contain advertiser purchase performance data. That is, there is a case where the purchase record data of a certain advertiser is accumulated in the second purchase record DB 6 but the purchase record data of the advertiser is not accumulated in the first purchase record DB 5 .

Alternatively, it is conceivable that purchase record data for a certain advertising space is accumulated in one of the first purchase record DB 5 and the second purchase record DB 6, but not accumulated in the other. In addition, the second purchase record DB 6 may contain purchase record data aggregated over a longer period than the first purchase record DB 5 .

In this way, the purchase performance data of the first purchase performance DB 5 and the purchase performance data of the second purchase performance DB 6 may differ in nature.
Therefore, as in the above embodiment, the purchase quantity (first purchase quantity) based on the purchase record data of the first purchase record DB 5 and the purchase quantity (second purchase quantity) based on the purchase record data of the second purchase record DB 6 It is conceivable that the purchase amount) and the purchase amount may be treated equally with one-to-one weighting, but there is a possibility that a more accurate recommendation can be made by integrating both purchase amounts with relatively different weights. That is, by weighting, when the first purchase amount is biased with respect to the second purchase amount, the bias can be corrected, and the ad space recommendation accuracy can be improved.

Note that in Equation 1 above, weighting is performed by uniformly multiplying all components of the second purchase amount matrix by the parameter α. However, the weighting method is not limited to this. For example, the first purchase amount matrix and the second purchase amount matrix may be weighted by multiplying each component of the second purchase amount matrix by a different parameter. In other words, in the purchase amount matrix, different weights may be applied depending on advertisers and advertising spaces. Also, weighting may be performed by any other method.

Further, in the above example, the purchase amount matrix having advertiser IDs as rows was illustrated as the first purchase amount matrix and the second purchase amount matrix, but the purchase amount matrix is not limited to this.
For example, the first and second purchase quantity matrices are [4. Other embodiments], such as a matrix whose rows correspond to the IDs of the advertiser's industry or product service category, whose columns correspond to the advertising space IDs, and whose component is the average purchase amount, etc. may be Then, these purchase amount matrices may be weighted. Such a purchase amount matrix also corresponds to the first and second purchase amounts. That is, the first (second) purchase amount is one or more advertising spaces of one or more advertisers specified from the purchase record data acquired from the first purchase record DB 5 (second purchase record DB 6). It may be an amount based on the purchase amount for

(11) In the above embodiment, a package product, which is a product in which a plurality of advertising spaces are grouped together, may be recommended to the user.
In this case, for example, a packaged product in which a plurality of advertising spaces purchased for the same campaign are bundled together may be recommended to the user.

Specifically, for example, the control unit 13 may acquire purchase record data for this type of package product as follows, and recommend the package product to the user based on the acquired purchase record data.

That is, as described above, for multiple ad slots traded using the ad inventory trading platform, system 100 can determine whether they were purchased for the same campaign. For this reason, the first purchase record DB 5 can store the purchase record of package products composed of a plurality of advertising spaces purchased for the same campaign.

On the other hand, for multiple ad spaces traded without using the ad space trading platform, the system 100 may not be able to determine whether they were purchased for the same campaign. For this reason, as described above, multiple advertising spaces with the same advertiser, advertisement placement period, advertised product or service, brand, etc. specified from each advertisement placed using those advertising spaces determined to be a plurality of ad slots purchased for the campaign. As a result, it is possible to generate the purchase record data of package products based on the purchase record data accumulated in the second purchase record DB 6 .

Then, using the package product purchase record data accumulated in both DBs 5 and 6, the package product can be handled in the same manner as other products (that is, products consisting of one advertising space) in the purchase amount matrix Y1. can. Package products may be recommended to the user using such a purchase amount matrix Y1.

(12) A plurality of functions possessed by one component in the above embodiment may be realized by a plurality of components, or a function possessed by one component may be realized by a plurality of components. . Also, a plurality of functions possessed by a plurality of components may be realized by a single component, or a function realized by a plurality of components may be realized by a single component. Also, part of the configuration of the above embodiment may be omitted. Moreover, at least part of the configuration of the above embodiment may be added or replaced with respect to the configuration of the other above embodiment.

(13) In addition to the control unit 13 of the advertising space trading device 1 described above, the advertising space trading system 100 having the advertising space trading device 1 as a component, a program for causing a computer to function as the control unit 13, and the program are stored. The present disclosure can also be implemented in various forms such as a non-transitional substantive storage medium such as a semiconductor memory, an advertising space trading method, and the like.

1... Advertising space trading device, 2... Provider side terminal, 3... User side terminal,
4... Advertising space information database, 5... First purchase record database,
6... Second purchase record database, 100... Advertising space transaction system.

Claims (11)

An advertising space trading system for providing a platform for trading advertising space on a network, comprising:
a publishing unit configured to publish information of a plurality of advertising slots related to reserved advertisements on the network;
an advertiser acquisition unit configured to acquire advertiser information, which is information about an advertiser, from a user;
A first purchase record database that accumulates purchase record data of the advertising space traded using the platform, and a second purchase record database that accumulates purchase record data of the advertising space traded without using the platform. a purchase record acquisition unit configured to acquire the purchase record data according to the advertiser information from each of a database;
a recommendation unit configured to recommend the advertising space to a user based on the purchase record data acquired from each of the first purchase record database and the second purchase record database;
an inventory trading system comprising: The advertising space trading system according to claim 1,
further comprising a campaign acquisition unit configured to acquire campaign information, which is information about the advertising campaign, from the user;
The ad space trading system, wherein the recommendation unit recommends the ad space to the user based on the campaign information. The advertising space trading system according to claim 1 or claim 2,
The advertiser information includes information on the advertiser's business type or product or service category,
The purchase record acquisition unit determines, from each of the first purchase record database and the second purchase record database, the type of business or the category of the advertiser who purchased the advertising space specified by the advertiser information. An advertising space trading system that acquires the purchase performance data of the advertising space that is the same as the business type or the category. The advertising space trading system according to any one of claims 1 to 3,
The second purchase record database accumulates the purchase record data of the advertising space identified from the already posted advertisements as the purchase record data of the advertising space traded without using the platform. ad space trading system. The advertising space trading system according to any one of claims 1 to 4,
further comprising a purchase candidate acquisition unit configured to acquire information on a purchase candidate frame that is the advertising space of the purchase candidate selected by the user;
The advertising space trading system, wherein the recommendation unit recommends the advertising space to the user based on information on the candidate purchase space. The advertising space trading system according to claim 5,
The advertising space trading system, wherein the recommendation unit preferentially recommends to the user the advertising space that is more likely to be purchased together with the purchase candidate frame in the same advertising campaign. The advertising space trading system according to claim 5 or claim 6,
The advertising space trading system, wherein the recommendation unit recommends the advertising space having the same or similar attributes as the purchase candidate frame to the user. The advertising space trading system according to any one of claims 1 to 7,
The recommendation unit selects the advertising space purchased in the past by the specific advertiser, or the An advertising space trading system for preferentially recommending the advertising space selected as a purchase candidate for an advertising campaign of a specific advertiser, and the advertising space having a higher probability of being purchased together. The advertising space trading system according to any one of claims 1 to 8,
further comprising a weighting unit configured to weight the first purchase quantity and the second purchase quantity;
The first purchase amount is the purchase amount for the one or more advertising slots of one or more advertisers, which is specified based on the purchase record data acquired from the first purchase record database, or the purchase is a quantity based quantity,
The second purchase amount is the purchase amount for the one or more advertising slots of one or more advertisers, which is specified based on the purchase record data acquired from the second purchase record database, or the purchase is a quantity based quantity,
The advertisement space trading system, wherein the recommendation unit recommends the advertisement space to the user based on the weighted first purchase amount and the second purchase amount. An advertising space trading method for providing a platform for trading advertising space on a network among computer operating methods, comprising :
Publishing information on a plurality of advertising slots related to reserved advertisements on the network;
Acquiring advertiser information, which is information about the advertiser, from the user;
A first purchase record database that accumulates purchase record data of the advertising space traded using the platform, and a second purchase record database that accumulates purchase record data of the advertising space traded without using the platform. obtaining the purchase performance data corresponding to the advertiser information from each of a database;
recommending the advertising space to a user based on the purchase record data acquired from each of the first purchase record database and the second purchase record database;
Inventory trading method, including; A computer program for causing a computer to function as an advertising space trading system for providing a platform for trading advertising space on a network,
a publishing unit configured to publish information of a plurality of advertising slots related to reserved advertisements on the network;
an advertiser acquisition unit configured to acquire advertiser information, which is information about an advertiser, from a user;
A first purchase record database that accumulates purchase record data of the advertising space traded using the platform, and a second purchase record database that accumulates purchase record data of the advertising space traded without using the platform. a purchase record acquisition unit configured to acquire the purchase record data according to the advertiser information from each of a database;
a recommendation unit configured to recommend the advertising space to a user based on the purchase record data acquired from each of the first purchase record database and the second purchase record database;
A computer program that causes a computer to function as a

Images