KR20210009128A - Method and system for advertisement exposure based on advertising effectiveness learning - Google Patents

Abstract

The present invention relates to a method for exposing an advertisement based on advertising efficiency. The method comprises the steps of: allocating the number of exposures of a specific advertisement for a plurality of time sections based on a preset distribution criterion, and exposing the specific advertisement according to the allocation; collecting an advertisement click event for the specific advertisement; monitoring whether a preset action is performed in connection with the advertisement click event while a session for the advertisement click event is activated; and calculating advertising efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result.

Description

Ad exposure method based on ad efficiency learning {METHOD AND SYSTEM FOR ADVERTISEMENT EXPOSURE BASED ON ADVERTISING EFFECTIVENESS LEARNING}

The present invention relates to an advertisement exposure method in consideration of advertisement efficiency.

As technology advances, online advertising in various media is increasing.

Such online advertisements are exposed through various devices such as PCs, laptops, tablets, and smartphones, and additional actions (eg, purchases) of consumers related to the advertiser's products or services are performed through the exposed advertisements.

In the case of such an online advertisement, the advertising cost varies based on the number of impressions of the advertisement or the number of clicks on the exposed advertisement. Therefore, it is the advertiser and the advertisement that expose the advertisement during the most effective time period within a limited budget. This is a very important issue for all agencies. Accordingly, in the field of online advertisement technology, active research is being conducted on various methods capable of maximizing advertisement efficiency.

An object of the present invention is to analyze an advertisement exposure time period that can maximize advertisement efficiency through learning about advertisements that have been executed in the past, and use the analyzed results to increase the efficiency of newly published advertisements. It is to provide an advertisement exposure method and system based on efficiency learning.

Another object of the present invention is to provide an advertisement exposure method and system based on advertisement efficiency learning that enables advertisements to be exposed in advertisement time periods with high actual advertisement efficiency by tracking the advertisement exposure time section in which an effective action has occurred. will be.

The advertisement exposure method based on advertisement efficiency learning according to the present invention includes: allocating the number of times of exposure of a specific advertisement for a plurality of time intervals based on a preset distribution criterion, and exposing the specific advertisement according to the allocation; Collecting an advertisement click event for the specific advertisement; Monitoring whether a preset action has been performed in connection with the advertisement click event while the session for the advertisement click event is activated; Calculating advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result; And updating the preset distribution criteria using the calculated advertisement efficiency data, and increasing the efficiency of the specific advertisement based on the updated distribution criteria for the time interval in which the advertisement click event occurs. It characterized in that it comprises the step of adjusting the number of exposures.

In an embodiment, in the collecting step, collecting a plurality of advertisement click events for the specific advertisement exposed in the plurality of time intervals, and in the monitoring step, the group among the plurality of advertisement click events It characterized in that it detects the advertisement click event connected to the set action.

In an embodiment, in the step of calculating the advertisement efficiency data, a time section to which the detected advertisement click event belongs is extracted from among the plurality of time sections, and the preset action is processed as an action occurring in the extracted time section. Characterized in that.

In an embodiment, the advertisement efficiency data is calculated using the number of advertisement click events in the extracted time interval and the number of actions that occur in connection with the extracted time interval.

In an embodiment, the distribution criterion is updated so that the number of exposures of the specific advertisement increases in the extracted time interval based on the processing of the preset action as an action occurring in the extracted time interval. .

In an embodiment, the preset action and the detected advertisement click event are included in the same session.

According to the present invention, an advertisement exposure system based on advertisement efficiency learning for improving advertisement efficiency of a specific advertisement exposed to a predetermined distribution criterion comprises: an advertisement click event collection unit configured to collect an advertisement click event for the specific advertisement; A monitoring unit that monitors whether a preset action is performed in connection with the advertisement click event while the session for the advertisement click event is activated; An advertisement efficiency data calculation unit that calculates advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result; And updating the preset distribution criteria using the calculated advertisement efficiency data, and increasing the efficiency of the specific advertisement based on the updated distribution criteria for the time interval in which the advertisement click event occurs. It characterized in that it comprises a control unit for adjusting the number of exposures.

As described above, in the advertisement exposure method and system based on advertisement efficiency learning according to the present invention, an advertisement is exposed for a plurality of advertisement exposure time intervals, and an additional action is performed in connection with an advertisement click event for the exposed advertisement. In this case, it is possible to track the advertisement exposure time interval associated with the corresponding action.

Furthermore, in the advertisement exposure method and system based on advertisement efficiency learning according to the present invention, when calculating the advertisement efficiency for each of a plurality of advertisement exposure time intervals, the additional action described above is not the time interval in which the additional action occurs, but the additional action It can be calculated as the advertisement efficiency of the time interval in which the advertisement click event associated with the action occurs. Accordingly, by allowing more advertisements to be exposed in the time period of the advertisement click event in which the actual additional action is performed, it is possible to induce more additional actions to occur in the time period in which advertisement efficiency is high.

1 is a block diagram illustrating an advertisement exposure system based on advertisement efficiency learning according to the present invention.
2 is a flowchart illustrating an advertisement exposure method based on advertisement efficiency learning according to the present invention.
3 and 4 are conceptual diagrams for explaining advertisement efficiency in the present invention.
5 to 8 are conceptual diagrams for explaining a method of calculating advertisement efficiency in the present invention.

Hereinafter, exemplary embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings, but the same reference numerals are assigned to the same or similar components regardless of the reference numerals, and redundant descriptions thereof will be omitted. The suffixes "module" and "unit" for components used in the following description are given or used interchangeably in consideration of only the ease of preparation of the specification, and do not have meanings or roles that are distinguished from each other by themselves. In addition, in describing the embodiments disclosed in the present specification, when it is determined that a detailed description of related known technologies may obscure the subject matter of the embodiments disclosed in the present specification, the detailed description thereof will be omitted. In addition, the accompanying drawings are for easy understanding of the embodiments disclosed in the present specification, and the technical idea disclosed in the present specification is not limited by the accompanying drawings, and all modifications included in the spirit and scope of the present invention It should be understood to include equivalents or substitutes.

Terms including ordinal numbers, such as first and second, may be used to describe various elements, but the elements are not limited by the terms. These terms are used only for the purpose of distinguishing one component from another component.

When a component is referred to as being "connected" or "connected" to another component, it is understood that it may be directly connected or connected to the other component, but other components may exist in the middle. Should be. On the other hand, when a component is referred to as being "directly connected" or "directly connected" to another component, it should be understood that there is no other component in the middle.

Singular expressions include plural expressions unless the context clearly indicates otherwise.

In the present application, terms such as "comprises" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof, does not preclude in advance.

The present invention relates to an advertisement system and method based on advertisement efficiency learning for increasing the efficiency of advertisement, in order to increase the efficiency of advertisements exposed through various media (eg, PC, laptop, tablet, smartphone), A method of learning data related to previously exposed advertisements will be described.

In this specification, for convenience of description, a medium through which advertisements are exposed will be collectively referred to as "terminal". Terminals described in the present specification include PCs, mobile phones, smart phones, laptop computers, digital broadcasting terminals, personal digital assistants (PDAs), portable multimedia players (PMPs), navigation, as described above. Slate PC, tablet PC, ultrabook, wearable device (e.g., smartwatch, smart glass), head mounted display (HMD) )), etc.

Meanwhile, the advertisement described in the present invention is included in at least a part of an application installed on the terminal or a web page output from the terminal and is exposed. Here, "the advertisement is exposed" means that the advertisement is displayed in a visual form on the terminal.

In performing advertisements on the terminal, there are various ways of executing advertisement expenses, as an example, CPM　 method (Cost Per Mille), which sets costs based on advertisement exposure (or number of exposures), CPP or CPT method based on period (Cost Per Time or Period), CPC 　 method (Cost Per Click) based on click (or selection) for exposed light, one-time conversion (for exposed advertisements) There is a CPA　 method (Cost Per Action or Acquisition) in which an advertisement fee is set based on a click and conversion to a preset action (eg, purchase, app installation, etc.) through click.

As described above, there are various ways of executing advertisement expenses, and advertisers wish to obtain maximum efficiency with a limited budget. Methods of calculating the efficiency of advertisements may be very diverse, and as an example, when an action desired by an advertiser occurs due to an exposed advertisement, it can be said that the advertisement efficiency is good.

Here, the advertiser's desired action means, for example, that the advertiser's product or service has been purchased through an exposed advertisement, an app (or application) has been installed, or that a member registration or consultation request has been made. It can mean, and the types of actions can be very diverse depending on the advertiser's request.

On the other hand, when an advertisement is exposed on a terminal, a certain time range may be set, such as a day unit or a week unit, to set the number of times the advertisement is exposed or the maximum number of clicks of the exposed advertisement. .

For example, if an advertisement is set to be exposed to a maximum of 2500 times a day, within 1 day, the advertisement may be exposed up to 2500 times, and beyond the number of times, the advertisement may no longer be exposed to the terminal.

For another example, when a maximum of 2,500 clicks per day are allowed for a given advertisement, when 2500 clicks are made on the advertisement within 1 day, the exposure is limited so that no more advertisements are clicked. This is to allow advertisements to be executed within a limited budget.

On the other hand, in the case of setting the number of times of exposure or clickable number of clicks so that the advertisement is exposed within a certain time range, the corresponding time range may have a plurality of time periods.

For example, if the number of exposures is set on a daily basis, it may have a plurality of time periods divided by a preset reference from 0 to 24 hours. Here, the preset time interval may be 24 time intervals, the first time zone (0 to 1:00), the second time zone (1 to 2:00), the third time zone (2:00 to 3:00), ?? , It can be divided into the 24th time zone (23 o'clock to 24 o'clock).

In this way, for a plurality of time intervals, the number of possible exposures or the number of clicks may be allocated respectively (hereinafter, for convenience of explanation, “the number of possible advertisements” and “the number of possible clicks of advertisements” are collectively referred to Therefore, the present invention will be referred to as “number of exposures.” Therefore, the present invention can be used to set the number of possible advertisements or the number of clicks available for advertisements).

Meanwhile, the advertisement efficiency is not the same for each of a plurality of time periods, and may have different advertisement efficiency for each of a plurality of time periods. Such advertising efficiency may vary depending on various variables such as the type of advertisement, the target of the advertisement, and the advertisement exposure time.

Therefore, it is important to allow advertisements to be exposed during times when advertisement efficiency is high. The present invention relates to a method of learning the efficiency of an advertisement that has already been exposed and reflecting it in the number of exposures of an advertisement to be newly executed. Hereinafter, together with the accompanying drawings, a more detailed look at.

Referring to FIG. 1, FIG. 1 is a block diagram illustrating an advertisement exposure system based on advertisement efficiency learning according to the present invention, and FIG. 2 is a flowchart illustrating an advertisement exposure method based on advertisement efficiency learning according to the present invention. Further, FIGS. 3 and 4 are conceptual diagrams for explaining advertisement efficiency in the present invention, and FIGS. 5 to 8 are conceptual diagrams for explaining a method of calculating advertisement efficiency in the present invention.

As shown in FIG. 1, the advertisement exposure system 100 based on advertisement efficiency learning according to the present invention includes an advertisement exposure distribution unit 110, an advertisement click event collection unit 120, a monitoring unit 130, and a communication unit ( 140), may include at least one of the data storage unit 150, the advertisement efficiency data calculation unit 160, and the control unit 170.

In FIG. 1, for convenience of explanation, the advertisement exposure system 100 based on advertisement efficiency learning according to the present invention is divided into a plurality of components. In some cases, the role of the plurality of components is a control unit 170 ) Can be performed.

Hereinafter, a method of controlling advertisement exposure through learning about an advertisement to be exposed, and hereinafter, a "specific advertisement" will be described. Here, the specific advertisement is an advertisement for which an advertisement is requested by an advertiser, and may be, for example, an advertisement for "mall A".

The advertisement exposure distribution unit 110 allocates the number of times of exposure of a specific advertisement for a plurality of time intervals based on a preset distribution standard.

Here, the plurality of time intervals are time intervals within a preset time range in which the advertisement is exposed as many as the preset number of exposures, and as described above, in the case of a daily exposure number, a preset standard from 0 to 24 hours. It may have a plurality of time intervals divided into. Here, the preset time interval may be 24 time intervals, the first time zone (0 to 1:00), the second time zone (1 to 2:00), the third time zone (2:00 to 3:00), ?? , It can be divided into the 24th time zone (23 o'clock to 24 o'clock).

Hereinafter, for convenience of explanation, it is assumed that a preset time range is assumed to be 1 day, and a plurality of time intervals are assumed to be time intervals consisting of a first time zone to a 24th time zone.

A preset allocation criterion for the number of exposures to a specific advertisement may be initially set by an advertisement publisher or programming.

If there is no initial distribution standard for the number of exposures, the controller 170 may allocate the number of exposures of the specific advertisement for each of the plurality of time intervals based on the advertisement distribution standard of a category similar to the specific advertisement.

Meanwhile, the controller 170 may control the communication unit 140 so that the specific advertisement is exposed to the terminal according to a preset distribution standard. The communication unit 140 may transmit information on the specific advertisement to an exposure target terminal under the control of the controller 170. Information on a specific advertisement may be pre-stored in the data storage unit 150 and exist.

The advertisement click event collection unit 120 is configured to collect an advertisement click event for the exposed specific advertisement when a specific advertisement is exposed for a plurality of time intervals based on a preset distribution criterion.

Here, the advertisement click event is an event that occurs when a click (or selection) for a corresponding advertisement is made in a terminal to which the specific advertisement is exposed.

In the present invention, "collecting an advertisement click event" includes a case in which information (eg, log information, etc.) indicating that the exposed specific advertisement has been clicked is received from an external server.

The monitoring unit 130 monitors whether an advertisement for a specific advertisement is clicked, and a preset action is performed in connection with the click.

Here, the preset action is an action of a consumer in connection with an exposed advertisement, for example, a purchase action associated with a specific advertisement, an app (or application) installation action, a member sign-up action, a consultation request action, etc. It can be very diverse.

When the preset action is performed, the monitoring unit 130 detects whether an advertisement click event having the same session as a session associated with the preset action exists. In addition, through such detection, the monitoring unit 130 may monitor the existence of an action performed in connection with the advertisement click event.

When the advertisement click event collection unit 120 collects a plurality of advertisement click events for the specific advertisement exposed in a plurality of time intervals, the monitoring unit 130 responds to the preset action among the plurality of advertisement click events. Detects a linked ad click event.

When a click is made to the specific advertisement while the specific advertisement is exposed, a session for the specific advertisement associated with the click is activated.

The monitoring unit 130 is configured to detect an action occurring in a session activated by a click event for the specific advertisement.

In addition, when the preset action is performed while the session is activated, in the present invention, the preset action is processed as an action by a specific advertisement associated with the activated session.

In this case, a time interval in which an advertisement click event for a specific advertisement occurs and a time interval in which a preset action occurs may be the same or different.

If the time interval in which the advertisement click event for a specific advertisement occurs and the time interval in which the preset action occurs are different, in the present invention, the advertisement click event for the specific advertisement occurs, compared to the time interval in which the preset action occurs. You can prioritize time intervals.

The control unit 170, by the monitoring unit 130, the time interval in which the preset action is detected and the preset action occurs, and a time interval in which the advertisement click event associated with the preset action occurs. In this case, the preset action may be processed as an action in a time interval in which the advertisement click event occurs.

For example, suppose that an ad click event occurs in the 7-8 hour period and a preset action occurs in the 15-16 hour period. In this case, by the monitoring unit 130, if the action occurring in the 15-16 time zone is originated from a session associated with the advertisement click event occurring in the 7-8 time zone, the controller 170 performs the preset action. , Is processed as originating from the advertisement click event.

In this case, the advertising efficiency may be calculated higher in the 7 to 8 hour interval.

For this processing, it is premised that the preset action is included in the same session as the session of the advertisement click event.

Meanwhile, the session information for such a preset action may be monitored by the system 100 according to the present invention or may be included in information transmitted from an external server through the communication unit 140.

In this case, the monitoring unit 130 extracts information on a preset action and a session connected to the action from information transmitted from an external server, and detects an advertisement click event associated with the extracted session. In addition, a time interval in which the detected advertisement click event occurs may be extracted.

The advertisement efficiency data calculation unit 160 calculates advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result.

Here, the advertisement efficiency may be calculated as “the number of times a preset action has occurred (or the number of action events)/the number of advertisement click events (or the number of click events)”, and the advertisement efficiency may also be referred to as conversion efficiency.

The advertisement efficiency data can be obtained for each of a plurality of time intervals in which the advertisement is exposed, and can be calculated using the number of advertisement click events in each time interval and the number of actions that occur in connection with the extracted time interval. have.

At this time, the control unit 170 connects the action occurring in the different time interval to the same session even if the time interval in which the advertisement click event occurs and the time interval in which the action occurs are different. It is processed as an action of an advertisement click event) included in the section. Therefore, according to the present invention, it is possible to extract a time section containing an advertisement click event that caused an actual action, and improve advertisement efficiency by calculating advertisement efficiency data so that the time section is reflected as a time section with high advertisement efficiency. I can make it.

Meanwhile, when advertisement efficiency data is calculated as above, the control unit 170 may update a preset distribution criterion that sets the number of advertisement exposures for a plurality of time intervals using the calculated advertisement efficiency data.

The distribution criterion is updated so that the number of exposures (or the number of clickables (or the maximum allowed number of clicks)) increases in a time period in which an advertisement click event that caused an actual action is included.

Conversely, for a time period in which the frequency of occurrence of an action is low, the preset allocation criterion may be updated so that the number of exposures (or the number of clicks possible (or the maximum number of allowed clicks)) decreases.

In addition, the control unit 170 exposes the specific advertisement based on the updated distribution criterion so as to increase the efficiency of the specific advertisement in a time interval in which an advertisement click event connected to the action (or a preset action) occurs. You can adjust the number of times.

That is, even if the time period in which the advertisement click event occurs and the time period in which the action occurs are different, the control unit 170 detects an action connected to the same session as the advertisement click event, thereby determining the detected action. It can be processed as an action that occurred in a time interval. Through this, in the present invention, by detecting the advertisement exposure time zone that contributed to the actual action, and re-adjusting the allocated number of exposures or the number of clicks (or the maximum allowed clicks) for each advertisement exposure time interval, based on the detection result, Advertising efficiency can be increased.

As described above, in the present invention, through learning to detect the time section of the advertisement click event of the same session as the session associated with the action generated through the previously exposed advertisement, and reflect the detected result in the advertisement efficiency data, the corresponding advertisement The allocation criteria for can be updated continuously.

Based on the above description, an advertisement exposure method of the advertisement exposure system 100 based on advertisement efficiency learning of the present invention will be described in more detail with reference to FIG. 2.

As shown in FIG. 2, according to the advertisement exposure method based on advertisement efficiency learning according to the present invention, first, a process of exposing a specific advertisement is performed based on a preset distribution criterion (S210).

The exposure of the specific advertisement is performed based on the number of allocations distributed through the advertisement exposure distribution unit 110, and the specific advertisement may be exposed by the control unit 170.

Here, the preset distribution criterion is, for each of a plurality of time periods, the number of exposures (eg, the number of advertisement exposures or the maximum allowed number of clicks) is set, and the control unit 170 is configured according to the preset distribution criterion. , You can display an advertisement, or limit the number of clicks of an advertisement.

A plurality of time intervals means that, for example, if the number of exposures is set in units of days, it is divided according to a preset standard from 0 to 24 hours, where there are 24 time intervals based on daily units. I can. The 24 time zones are in the 1st time zone (0~1pm), the 2nd time zone (1~2pm), the 3rd time zone (2pm~3pm), ??, and the 24th time zone (23:00~24:00). Can be distinguished.

As shown in FIG. 5 and shown, the number of allocated exposures may be set for each of a plurality of time periods according to a preset distribution criterion. In FIGS. 5 to 8, for explanation, it is assumed that the number of exposures to 1500 times is limited.

The "allocated number of exposures" of FIG. 5 may mean the number of exposures of an advertisement or the maximum number of clicks of an advertisement, and may be interpreted as an appropriate meaning in some cases.

As described above, when there is an advertisement exposure request for a specific advertisement, when pre-learned data does not exist, the control unit 170 can allocate the number of exposures arbitrarily or based on the advertisement distribution criteria of a similar category. .

In this way, when exposure to a specific advertisement is made according to the allocation, a process of collecting an advertisement click event for the specific advertisement is performed (S220).

Here, the advertisement click event is an event that occurs when a click (or selection) for a corresponding advertisement is made in a terminal to which the specific advertisement is exposed.

The advertisement click event may be collected by the advertisement click event collection unit 120 described above.

In the present invention, "collecting an advertisement click event" includes a case in which information (eg, log information, etc.) indicating that the exposed specific advertisement has been clicked is received from an external server.

As shown in FIG. 5, advertisement click events may be collected for each exposure time period. For example, in the "0" time zone, 30 advertisement click events may occur, and in the "1" time zone, 20 advertisement click events may occur.

As in step S220, when the advertisement click event is collected, a process of monitoring whether a preset action is performed in connection with the collected advertisement click event is performed (S230).

As described above, the preset action is a consumer's action in connection with the exposed advertisement, for example, a purchase action associated with a specific advertisement, an app (or application) installation action, a member sign-up action, a consultation request action Etc can be very diverse.

More specifically, when the preset action is performed, the monitoring unit 130 detects whether an advertisement click event having the same session as the session associated with the preset action exists. In addition, through such detection, the monitoring unit 130 may monitor the existence of an action performed in connection with the advertisement click event.

For example, as shown in FIG. 5, i) 30 click events occur during the exposure time "0", 3 action events (or preset actions) occur, and 20 times the exposure time "1" Assume that a conference click event occurs and two action events occur.

The number of advertisement click events and the number of action events shown in FIG. 5 are information counted based on each time period (or exposure time period). Therefore, the relationship between the click event and the action event is not reflected.

Accordingly, in the present invention, the exposure time interval in which the action event occurs may be adjusted by reflecting the relationship between the click event and the action event.

When the advertisement click event collection unit 120 collects a plurality of advertisement click events for the specific advertisement exposed in a plurality of time intervals, the monitoring unit 130 responds to the preset action among the plurality of advertisement click events. Detects a linked ad click event.

When a click is made to the specific advertisement while the specific advertisement is exposed, a session for the specific advertisement associated with the click is activated.

The monitoring unit 130 is configured to detect an action occurring in a session activated by a click event for the specific advertisement.

In addition, when the preset action is performed while the session is activated, in the present invention, the preset action is processed as an action by a specific advertisement associated with the activated session.

In this case, a time interval in which an advertisement click event for a specific advertisement occurs and a time interval in which a preset action occurs may be the same or different.

For example, referring to FIG. 5, it is assumed that 10 ad click events and 10 action events occur in the exposure time "8", and 20 click events and action events occur in the exposure time "9". . In this case, based on the exposure time, it can be seen that the advertisement efficiency (a) in the "8" time period is 100%, and the advertisement efficiency (a) in the "9" exposure time period is very high as 25%.

However, as a result of the monitoring, as shown in FIG. 6, if the session of the advertisement event associated with the action event in the "8" exposure time zone is connected to the advertisement click event in the "0" time zone, not the "8" time zone, As a result, the exposure time zone in which advertising efficiency should be high will be in the "0" time zone. Accordingly, in the present invention, as shown in Figs. 5 and 6, the advertisement efficiency (a) in the "0" time period is corrected to 43.3%, not 10%. As shown in FIG. 7, in reality, in the "0" time zone, an action associated with an advertisement event occurring in a corresponding time interval is extracted, and the number of extracted actions is processed as occurring in a time interval of the associated advertising event ( The number of action events before improvement in the "0" time zone: 3 times, the number of action events after improvement: 13 times/ The number of action events before improvement in the "8" time zone: 10 times, the number of action events after improvement: 0 times).

As described above, when the time interval in which the advertisement click event for a specific advertisement occurs and the time interval in which the preset action occurs are different, in the present invention, the advertisement click event for the specific advertisement is more than the time interval in which the preset action occurs. Priority can be given to the time interval that occurred. The control unit 170, by the monitoring unit 130, the time interval in which the preset action is detected and the preset action occurs, and a time interval in which the advertisement click event associated with the preset action occurs. In this case, the preset action may be processed as an action in a time interval in which the advertisement click event occurs.

Accordingly, in the present invention, by detecting an advertisement event associated with the performed action, a time period having high actual advertisement efficiency is extracted, and the number of advertisement exposures (or clickable number) is increased in the corresponding time period, thereby increasing advertisement efficiency. I can.

On the other hand, as in step S230, when monitoring is performed, a process of calculating advertisement efficiency data for a specific advertisement using the advertisement click event and the monitoring result is performed (S240).

The advertisement efficiency data calculation unit 160 calculates advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result.

Here, the advertisement efficiency may be calculated as "the number of times a preset action has occurred (or the number of action events) / the number of advertisement click events (or the number of click events)", as shown in FIG. 3(a). , Advertising efficiency may also be referred to as conversion efficiency.

The advertisement efficiency data can be obtained for each of a plurality of time intervals in which the advertisement is exposed, and can be calculated using the number of advertisement click events in each time interval and the number of actions that occur in connection with the extracted time interval. have.

In this way, when the advertisement efficiency data is calculated, the maximum efficiency distribution per click (per advertisement click event) may be derived, as shown in FIG. 4, based on the equation shown in FIG. 3B. This maximum efficiency distribution map reflects an action occurring in the same session as each advertisement event, and may be used to refer to an advertisement exposure section having a high actual efficiency.

Meanwhile, (c) of FIG. 3 is a cumulative advertisement efficiency distribution value, which is a value obtained by accumulating advertisement efficiency in each time interval, and the total is 1.

Meanwhile, the advertisement efficiency data calculated in consideration of the above-described method is compared with the advertisement efficiency data calculated based on different criteria to be examined.

First, in FIGS. 7 and 8, exposure ratios according to first to third advertisement efficiencies are shown, and the sum of each exposure ratio according to first to third advertisement efficiencies is 1.

The first advertisement efficiency means "the number of advertisement click events/the number of allocated exposures (or the number of clicks allocated and the maximum number of clicks)" and means the number of times the advertisement is clicked compared to the number of times the advertisement is exposed.

The second advertisement efficiency is "the number of action events (or actions) in the corresponding time interval (in Fig. 7, the number of action events before improvement)/the number of advertisement click events", and the correlation between the advertisement click event and the action (for example, , Whether or not they are included in the same session).

In addition, the third advertisement efficiency is "the number of action events linked to the collection time interval of the number of advertisement click events (number of mutually related action events, number of action events after improvement in FIG. 7) / number of advertisement click events", The correlation between events and actions (eg, whether they are included in the same session) is considered.

Looking at the exposure ratio according to the second and third advertisement effects in the "0" exposure time period of FIG. 7, the exposure ratio (0.3333) when the correlation between the advertisement click event and the action is considered, and the exposure ratio ( 0.0508) is very different.

In this way, when the advertisement efficiency data is calculated, a process of updating a preset distribution standard and adjusting the number of exposures of a specific advertisement is performed using the advertisement efficiency data (S250).

More specifically, the control unit 170 connects the action occurring in the different time interval to the same session, even if the time interval in which the advertisement click event occurs and the time interval in which the action occurs are different, if they are connected through the same session (however, It is handled as an action of an advertisement click event) included in another time period. Therefore, according to the present invention, it is possible to extract a time section containing an advertisement click event that caused an actual action, and improve advertisement efficiency by calculating advertisement efficiency data so that the time section is reflected as a time section with high advertisement efficiency. I can make it.

Meanwhile, when the advertisement efficiency data is calculated as above, the control unit 170 updates a preset distribution criterion that sets the number of advertisement exposures for a plurality of time intervals using the calculated advertisement efficiency data.

The distribution criterion is updated so that the number of exposures (or the number of clicks (or the maximum number of clicks)) is increased in the time period in which the advertisement click event that caused the actual action is included.

Conversely, for a time period in which the frequency of occurrence of an action is low, the preset allocation criterion may be updated so that the number of exposures (or the number of clicks possible (or the maximum number of allowed clicks)) decreases.

In addition, the control unit 170 exposes the specific advertisement based on the updated distribution criterion so as to increase the efficiency of the specific advertisement in a time interval in which an advertisement click event connected to the action (or a preset action) occurs. You can adjust the number of times.

That is, even if the time period in which the advertisement click event occurs and the time period in which the action occurs are different, the control unit 170 detects an action connected to the same session as the advertisement click event, thereby determining the detected action. It can be processed as an action that occurred in a time interval.

Through this, in the present invention, by detecting the advertisement exposure time zone that contributed to the actual action, and re-adjusting the number of exposures or the number of clicks (or the maximum allowed clicks) allocated for each advertisement exposure time interval based on the detection result, Advertising efficiency can be increased.

As described above, in the present invention, through learning to detect the time section of the advertisement click event of the same session as the session associated with the action generated through the previously exposed advertisement, and reflect the detected result in the advertisement efficiency data, the corresponding advertisement The allocation criteria for can be updated continuously.

Furthermore, the control unit 170 according to the present invention may continuously accumulate learning data by repeatedly performing the processes S210 to S250 described in FIG. 2.

More specifically, in step S210, the "preset distribution criterion", which is the exposure criterion of the first advertisement, is an advertisement exposure for the first advertisement performed in the past based on the time point at which the first advertisement is to be exposed, and the advertisement It can be continuously updated based on efficiency results.

For example, during the advertisement period of 10 days, in the advertisement on the 3rd day, the number of exposures to the first advertisement is adjusted based on the advertisement results performed on the 1st to 2nd days and the advertisement efficiency data therefor. In addition, in the advertisement of the 6th day, the number of exposures to the first advertisement may be adjusted based on the advertisement result performed on the 1st to the 5th day and advertisement efficiency data therefor. That is, the control unit 170 accumulates advertisement efficiency data for previously performed advertisements and uses the accumulated advertisement efficiency data when distributing the number of advertisement exposures.

Accordingly, the control unit 170 sets a "preset distribution criterion" for advertisement exposure on the 10th day by using advertisement efficiency data accumulated for 1 to 9 days.

This preset distribution criterion is updated by accumulating advertisement efficiency data for the advertisement that has been previously processed as the advertisement progresses. The control unit predicts the advertisement efficiency for the advertisement to be proceeded based on the updated pre-set distribution criterion, and sets the number of exposures based on the prediction.

Here, predicting the advertising efficiency may be understood as "setting the number of exposures to the advertisement to be promoted by reflecting the advertisement efficiency trend according to the advertisement efficiency data performed in the past".

That is, in the present invention, the number of advertisement exposures may be set by reflecting the advertisement efficiency trend of the advertisements exposed as the advertisement exposure period progresses, using advertisements exposed in the past and advertisement efficiency data therefor.

Therefore, according to the present invention, by continuously monitoring the reaction of users (for example, terminal users to which the advertisement is exposed) to the advertisement according to the progress of the exposure to the advertisement target advertisement, You can set the optimal number of exposures.

As described above, the advertisement exposure method based on advertisement efficiency learning according to the present invention calculates advertisement efficiency data by extracting an advertisement click event connected to the same session as a preset action during the advertisement exposure period, and continuously calculates the advertisement efficiency data. It is possible to predict an advertisement time period with relatively high advertisement efficiency by reflecting the trend of advertisement efficiency according to the progress of advertisement using the learning result.

On the other hand, looking at a specific example in which the distribution criterion is updated according to the advertisement exposure ratio included in the calculated advertisement data described above, the control unit 170, the number of exposures (or the number of clicks allocated, the maximum Clicks allowed) can be redistributed (corresponds to updates). As shown in FIG. 8, if about 76 exposures are allocated before improvement in the "0" exposure time zone based on the limited number of exposures of 1500, about 500 exposures may be allocated after improvement. In other words, in the present invention, advertising efficiency can be improved by allocating a greater number of exposures to a time period consisting of an actual action.

On the other hand, according to this criterion, after the number of exposures is allocated for each time period, if the number of exposures as many as the number of exposures allocated in the corresponding period is not made (or when no clicks are made), the controller 170 You can redistribute the number of remaining exposures to different time intervals.

In this case, the controller 180 may redistribute the number of remaining exposures to different time intervals based on various criteria. As an example, the control unit 170 is the number of remaining exposures in the next time interval immediately consecutive to the time interval of the remaining number of exposures. Can be assigned. As another example, the control unit 170 may reallocate the number of remaining exposures based on a time period in which advertisement efficiency is high among the remaining time periods.

On the other hand, for a specific advertisement, as in the above-described method, when there is no pre-learned advertisement efficiency data, the controller 170 may have a category similar to the specific advertisement based on the information stored in the data storage unit 150. Based on the learned advertisement efficiency data of the advertisement, the number of exposures can be allocated.

As described above, in the advertisement exposure method and system based on advertisement efficiency learning according to the present invention, an advertisement is exposed for a plurality of advertisement exposure time intervals, and an additional action is performed in connection with an advertisement click event for the exposed advertisement. In this case, it is possible to track the advertisement exposure time interval associated with the corresponding action.

Furthermore, in the advertisement exposure method and system based on advertisement efficiency learning according to the present invention, when calculating the advertisement efficiency for each of a plurality of advertisement exposure time intervals, the additional action described above is not the time interval in which the additional action occurs, but the additional action It can be calculated as the advertisement efficiency of the time interval in which the advertisement click event associated with the action occurs. Accordingly, by allowing more advertisements to be exposed in the time period of the advertisement click event in which the actual additional action is performed, it is possible to induce more additional actions to occur in the time period in which advertisement efficiency is high.

Claims (7)

Allocating the number of times of exposure of a specific advertisement for a plurality of time intervals based on a preset distribution criterion, and exposing the specific advertisement according to the allocation;
Collecting an advertisement click event for the specific advertisement;
Monitoring whether a preset action has been performed in connection with the advertisement click event while the session for the advertisement click event is activated;
Calculating advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result; And
The specific advertisement is exposed for a time interval in which the advertisement click event occurs based on the updated distribution criteria to update the preset distribution criteria using the calculated advertisement efficiency data and increase the efficiency of the specific advertisement Ad exposure method based on advertising efficiency learning, comprising the step of adjusting the number of times. The method of claim 1,
In the collecting step,
Collecting a plurality of advertisement click events for the specific advertisement exposed in the plurality of time intervals,
In the monitoring step,
An advertisement exposure method based on advertisement efficiency learning, characterized in that detecting an advertisement click event connected to the preset action among the plurality of advertisement click events. The method of claim 2,
In the step of calculating the advertisement efficiency data,
Extracting a time section to which the detected advertisement click event belongs among the plurality of time sections,
The advertisement exposure method based on advertisement efficiency learning, characterized in that processing the preset action as an action occurring in the extracted time interval. The method of claim 3,
The advertisement efficiency data is calculated using the number of advertisement click events in the extracted time interval and the number of actions that occur in connection with the extracted time interval. The method of claim 3,
Based on the processing of the preset action as an action occurring in the extracted time period, the distribution criterion is updated to increase the number of exposures of the specific advertisement in the extracted time period. How to display ads. The method of claim 2,
The advertisement exposure method based on advertisement efficiency learning, characterized in that the preset action and the detected advertisement click event are included in the same session. In the advertisement exposure system based on advertisement efficiency learning to improve the advertisement efficiency of a specific advertisement exposed to a preset distribution standard,
An advertisement click event collection unit that collects an advertisement click event for the specific advertisement;
A monitoring unit that monitors whether a preset action is performed in connection with the advertisement click event while the session for the advertisement click event is activated;
An advertisement efficiency data calculation unit that calculates advertisement efficiency data for the specific advertisement by using the advertisement click event for the specific advertisement and the monitoring result; And
The specific advertisement is exposed for a time interval in which the advertisement click event occurs based on the updated distribution criteria to update the preset distribution criteria using the calculated advertisement efficiency data and increase the efficiency of the specific advertisement An advertisement exposure system based on advertisement efficiency learning, comprising a control unit for adjusting the number of times.

Images