KR20150111027A - Apparatus and method for providing advertisements - Google Patents

Abstract

An apparatus for supplying advertisements to a user includes an advertisement management unit which manages a plurality of advertisements and the exposure and performance history of each advertisement, an advertisement performance index management unit which calculates a probability distribution of a performance index according to each advertisement based on the performance history and exposure of each advertisement, and an advertisement providing unit which receives an advertisement request from a user terminal, extracts the performance index of each advertisement based on the probability distribution of the performance index according to each advertisement, and selects the advertisement to be supplied to the user terminal based on the performance index of each advertisement.

Description

[0001] APPARATUS AND METHOD FOR PROVIDING ADVERTISEMENTS [0002]

The present invention relates to an apparatus and a method for providing an advertisement.

The user inputs a search word reflecting the interest in a portal site or the like, and receives search results related to the search term. At this time, a company that operates a portal site or the like exposes the advertisement requested by the advertiser to the search result through the advertisement providing device. Then, the operator and the advertiser select the advertisement, execute the advertisement, and determine the price according to the response information such as the click through rate (CTR).

Until now, the advertisement is executed based on the calculated performance index based on the advertisement execution information. Therefore, until the advertisement execution information is updated and the performance indicator is calculated again, the advertisement is executed based on the determined performance indicator. Also, since the performance indexes so far are calculated at a rate like the click rate per impression irrespective of the number of times of the advertisement exposure, the number of the advertisement exposure is not reflected in the advertisement execution. Therefore, it is difficult to evaluate the performance of the new advertisement only by the statistical calculated indexes because the number of exposures and the number of clicks are small. In such a case, performance indexes are predicted using performance indexes of other advertisements, but this is only a temporary measure, and it is difficult to view them as performance indicators of new advertisements.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an apparatus and a method for variably extracting an advertisement performance index based on a probability distribution reflecting the number of times of advertisement exposure and providing an advertisement based on the extracted performance index.

An advertisement providing apparatus for providing an advertisement to a user according to an embodiment of the present invention includes an advertisement management unit for managing a plurality of advertisements and an exposure and a performance history of each advertisement, The advertisement performance index management unit which calculates the probability distribution of the performance indicators separately and the advertisement reception requests from the user terminals and extracts the performance indicator values of each advertisement based on the probability distribution of the advertisement performance indicators, And an advertisement providing unit for selecting an advertisement to be provided to the user terminal as a basis.

The advertisement performance indicator management unit calculates an average and a standard deviation relating to the performance indicators of the respective advertisements based on the exposure and the performance history related to the performance indicators of the respective advertisements, The probability distribution of the performance index of the server can be calculated.

When the exposure and the performance history of a certain advertisement are updated among the plurality of advertisements, the advertisement performance indicator management unit updates the probability distribution of the performance indicator corresponding to the arbitrary advertisement.

Wherein the advertisement providing unit extracts a first performance index value from a probability distribution of a performance index corresponding to the first advertisement when receiving the first advertisement request and if the second advertisement request is received from the probability distribution of the performance index corresponding to the first advertisement, The second performance index value can be extracted from the distribution.

The performance measure may include at least one of a click rate per impression and a response rate per impression.

The probability distribution may include at least one of a normal distribution, a beta distribution, and a Poisson distribution.

According to another aspect of the present invention, there is provided a method of providing an advertisement to a user, the method comprising: managing an exposure and a performance history of each of a plurality of advertisements; A step of calculating a probability distribution of the performance indexes, a step of extracting performance index values of each advertisement based on the probability distribution of the performance indexes for each advertisement, and a step of selecting an advertisement to be provided to the user terminal based on the measurement index values of the respective advertisements .

The step of calculating the probability distribution of the performance index by the advertisement calculates the average and standard deviation of the performance index of each advertisement based on the exposure and the performance history related to the performance index of each advertisement, The probability distribution of the performance index can be calculated on a per-advertisement basis.

The probability distribution may include at least one of a normal distribution, a beta distribution, and a Poisson distribution.

The performance measure may include at least one of a click rate per impression and a response rate per impression.

The step of extracting the performance index value of each advertisement may extract the performance index value of each advertisement based on the probability distribution of the performance index by the advertisement when receiving the advertisement request from the user terminal.

Wherein the step of extracting the performance index values of each advertisement extracts a first performance index value from the probability distribution of the performance indexes corresponding to the first advertisement upon receiving the first advertisement request, The second performance index value can be extracted from the probability distribution of the performance index corresponding to the advertisement.

The advertisement providing method may further include updating a probability distribution of a performance index corresponding to the random advertisement when the exposure and the performance history of the advertisement are updated among the plurality of advertisements.

According to the embodiment of the present invention, the performance index of the advertisement can be predicted based on the exposure and response information of the advertisement, and the advertisement can be executed based on the predicted performance index. Therefore, according to the embodiment of the present invention, the exposure index and the response information of the advertisement are reflected in the probability distribution of the performance index, and the performance index value is extracted from the probability distribution, so that the performance index value in the reliable interval can be obtained.

According to the embodiment of the present invention, the performance index is extracted according to the probability distribution of the performance index per advertisement request, so that the performance index value is variable. Therefore, according to the embodiment of the present invention, it is possible to increase the advertisement revenue and the advertising effect by giving an opportunity to be executed even if the advertisement has a small number of responses such as exposure or click.

1 is a schematic diagram of an advertising system in accordance with an embodiment of the present invention.
FIGS. 2 to 4 are diagrams illustrating a probability distribution of a performance index according to an embodiment of the present invention.
5 is a flowchart of an advertisement providing method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification, when an element is referred to as "comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise. Also, the terms " part, "" module," and " module ", etc. in the specification mean a unit for processing at least one function or operation and may be implemented by hardware or software or a combination of hardware and software have.

FIG. 1 is a schematic diagram of an advertisement system according to an embodiment of the present invention, and FIGS. 2 to 4 are diagrams illustrating a probability distribution of a performance index according to an embodiment of the present invention.

Referring to FIG. 1, the advertisement system 10 includes an advertisement providing apparatus 100, an advertiser terminal 200, and a user terminal 300.

The advertisement providing apparatus 100 provides an advertisement of the advertiser terminal 200 to the user terminal 300 according to an advertisement service application of the advertiser terminal 200.

The advertiser terminal 200 is an apparatus in which an advertiser connects to and communicates with the advertisement providing apparatus 100 through a network. The advertiser terminal 200 includes a computer, a personal digital assistant (PDA), a mobile communication terminal, TV) can be used.

The user terminal 300 is a device through which a user connects to and communicates with the advertisement providing apparatus 100 through a network. Like the advertiser terminal 200, various communication apparatuses such as a computer, a personal digital assistant, a mobile communication terminal, .

The advertisement providing apparatus 100 includes an advertisement managing unit 110, an advertisement performance indicator managing unit 130, and an advertisement providing unit 150.

The advertisement management unit 110 manages a plurality of advertisements, and exposure and performance history of each advertisement. Exposure and performance history include exposure information and various information related to post-exposure performance. For example, exposure and performance history include number of exposures, number of clicks, number of actions (actions), and the like. Here, the response includes a response of the user in various forms to the advertisement, and may include, for example, mouse over, video playback, video extension, and the like. For example, the advertisement managing unit 110 manages the exposure and the performance history of the advertisements A and B as shown in Table 1.

Number of impressions Clicks Number of reactions (actions) Ad A 10000 8 20 Ad B 1000 One 5

The advertisement performance indicator management unit 130 calculates a probability distribution of the performance indicators for each advertisement based on the exposure and the performance history of each advertisement. The performance indicators can vary and can be, for example, click through rate (CTR), response rate per impression, and the like.

)를 구한다. The probability distribution of the performance index can be calculated by various statistical methods. For example, probability distributions can be expressed by various kinds of probability density functions. For example, a probability distribution may include various kinds of probability distributions, such as a normal distribution, a beta distribution, and a Poisson distribution, which are calculated on the basis of mean and variance. For example, when the advertisement performance indicator 130 calculates a probability distribution of a performance index using a normal distribution, the average (m) and the standard deviation (

).

)를 기초로 도 2와 같은 모양의 확률 분포를 구한다. 광고A와 광고B의 노출당 클릭률의 확률 분포는 표 2와 같은 평균과 표준 편차를 가지고, 도 2와 같은 확률 분포 모양을 가진다. 광고 성과 지표 관리부(130)는 노출당 반응률의 확률 분포를 구할 수 있고, 여기서는 노출당 클릭률(CTR)의 확률 분포를 기준으로 설명한다.The advertising performance indicator manager 130 obtains average (m) and standard deviation (

The probability distribution of the shape shown in FIG. 2 is obtained. The probability distribution of the click rate per advertisement of advertisement A and advertisement B has an average and standard deviation as shown in Table 2 and has a probability distribution shape as shown in FIG. The advertisement performance indicator management unit 130 can obtain the probability distribution of the response rate per impression, and explains the probability distribution of the click-through rate (CTR) per impression here.

Average Standard Deviation Ad A 0.008 (0.08%) 0.00028 Ad B 0.01 (0.1%) 0.001

When the exposure and the performance history of each advertisement are updated, the advertisement performance indicator management unit 130 also updates the probability distribution of the performance indicators. Referring to FIG. 3, as the advertisement is executed, the average may change, and the probability density near the average gradually increases. Therefore, as the advertisement is executed, the probability distribution becomes narrower and the probability density value becomes larger. That is, the probability that the measurement index value is extracted near the average becomes higher.

The advertisement providing unit 150 receives an advertisement request from the user terminal 300. The advertisement providing unit 150 delivers at least one advertisement stored in the advertisement managing unit 110 to the user terminal 300. At this time, the advertisement providing unit 150 determines which of the plurality of advertisements stored in the advertisement managing unit 110 is to be executed based on the advertisement execution criteria. The advertising execution standard may vary not only as a performance index but also as a price. Here, it is assumed that the advertisement is executed based on the performance index.

범위 내, 또는 2

범위 내의 성과 지표값을 추출하도록 추출 범위를 설정할 수 있다.The advertisement providing unit 150 executes the advertisement based on the performance indicators of each advertisement probabilistically extracted from the advertisement performance indicator management unit 130. That is, the advertisement provider 150 does not use the fixed performance index value but uses the performance index value extracted from the probability distribution of the performance index. Therefore, the measurement index value of each advertisement may be changed each time the user requests the advertisement. The advertisement providing unit 150 can set the extraction range in the probability distribution. For example, the advertisement providing unit 150 may generate 1

Within range, or 2

The extraction range can be set to extract the measurement index values within the range.

For example, referring to FIG. 4, the average CTR (m_a) of advertisement A is lower than the average CTR (m_b) of advertisement B. Since the average CTR (m_b) of the advertisement B is small in the number of exposures, the average CTR (m_b) may not yet be a reliable value. Nonetheless, if an ad is executed based only on the average CTR, the ad B with the highest average CTR is always executed before the ad A at all times.

On the other hand, the advertisement providing unit 150 executes the advertisement considering the dispersion due to the number of exposures. The advertisement provider 150 extracts an arbitrary performance index value from the probability distribution of the performance indexes of each advertisement, and compares the extracted performance index values to execute the advertisement.

Ad A has a high probability density near average CTR. Therefore, advertisement A has a high probability that the performance index value of the CTR per impression is extracted near the average CTR, and the advertisement B has a value that the average CTR is larger than the advertisement A but is far from the average CTR (for example, A value lower than the measurement index value) may be extracted. If the performance index value m_a1 extracted from the probability distribution of the advertisement A is larger than the performance index value m_b1 extracted from the probability distribution of the advertisement B, the advertisement providing unit 150 may execute the advertisement A first.

In this way, when the extracted value is used according to the probability distribution rather than using the fixed performance index value, the advertisement provider 150 may change the size of the performance index value. Therefore, the advertisement providing unit 150 may also execute an advertisement having a low average performance index value, thereby giving an opportunity to increase the performance.

5 is a flowchart of an advertisement providing method according to an embodiment of the present invention.

Referring to FIG. 5, the advertisement providing apparatus 100 calculates a probability distribution of a performance index for each advertisement based on the exposure and the performance history of each advertisement (S110). For example, exposure and performance history include number of exposures, number of clicks, number of reactions, and the like. Performance metrics can be clickthrough rate (CTR) per impression, response rate per impression, and so on.

The advertisement providing apparatus 100 receives an advertisement request from the user terminal 300 (S120).

The advertisement providing apparatus 100 extracts the performance indicator values for each advertisement in the probability distribution of the performance indicators of each advertisement (S130).

The advertisement providing apparatus 100 determines an advertisement to be provided to the user terminal 300 based on the measurement index value for each advertisement (S140).

For example, if advertisement C has an exposure count of 10000, a click count of 5, an average CTR of 0.0005, and advertisement D has an exposure count of 10,000, a click count of 6, and an average CTR of 0.0006, Ad D is always selected by the difference.

On the other hand, the advertisement providing apparatus 100 extracts the performance indicator values of each advertisement from the probability distribution of the performance indicators of the respective advertisements. Therefore, referring to Table 3, since the CTR is extracted in order to execute the advertisement, the CTR of each of Advertisements C and D is changed. In some cases, the CTR of the advertisement C may be greater than the CTR of the advertisement D. For example, the advertisement providing apparatus 100 may select the advertisement C for the secondary advertisement request. In this manner, when the advertisement providing apparatus 100 extracts the performance indicator values of the respective ads from the probability distributions of the performance indicators of the advertisements C and D for each advertisement request, and selects the advertisement based on the extracted performance indicator values, , Ad C and ad D will be executed at a ratio of c: d. For example, c may be 39, and d may be 61. That is, the advertisement providing apparatus 100 can execute the advertisement based on the performance index values extracted in real time when the advertisement is requested, rather than executing the advertisement based on the execution ratio of the advertisement calculated in advance.

Ad C Ad D Selection Result Primary 0.00043 0.00081 Ad D Secondary 0.00078 0.00052 Ad C Third 0.00042 0.00064 Ad D Fourth 0.00032 0.00061 Ad D 5th 0.00079 0.00126 Ad D

In this way, when the number of exposures, the number of clicks, the number of responses, etc. are not sufficient, the advertisement providing apparatus 100 can reduce an advertisement execution error generated by using the fixed performance index value. Also, instead of extracting the performance index values randomly between 0 and 1, the advertisement providing device 100 extracts the performance index values from the probability distribution of the performance indexes generated based on the exposure and the performance history of each advertisement. Therefore, the advertisement providing apparatus 100 mainly extracts a value near the average based on the probability distribution, but also extracts a value that is away from the average, and adjusts the result so that the advertisement is selected flexibly.

When the number of exposures, the number of clicks, and the number of responses are sufficient, the probability distribution of the performance index changes to a probability density of the average value close to one. In this case, the advertisement providing apparatus 100 may execute the advertisement based on the average performance index value.

The embodiments of the present invention described above are not implemented only by the apparatus and method, but may be implemented through a program for realizing the function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, It belongs to the scope of right.

Claims (13)

An advertisement providing apparatus for providing an advertisement to a user,
An advertisement management unit for managing a plurality of advertisements, exposure and performance history of each advertisement,
An advertisement performance indicator management unit for calculating a probability distribution of the performance indicators for each advertisement based on the exposure and the performance history of each advertisement,
An advertisement providing system for receiving an advertisement request from a user terminal, extracting a measurement index value of each advertisement based on a probability distribution of the advertisement measurement indexes, and selecting an advertisement to be provided to the user terminal based on the measurement index value of each advertisement
And an advertisement providing device. The method of claim 1,
The advertising performance indicator
The average and standard deviation of the performance indexes of the respective advertisements are calculated based on the exposure and performance histories related to the performance indicators of the respective advertisements. Based on the average and the standard deviation of each advertisement, the probability distribution To the advertisement server. The method of claim 1,
The advertising performance indicator
And updates a probability distribution of a performance index corresponding to the random advertisement when the exposure and the performance history of an arbitrary advertisement among the plurality of advertisements are updated. The method of claim 1,
The advertisement providing unit
The method according to claim 1 or 2, wherein when the first advertisement request is received, a first performance index value is extracted from a probability distribution of a performance index corresponding to the first advertisement, Wherein the value of the performance indicator is extracted. The method of claim 1,
The performance indicator
A click rate per impression, and a response rate per impression. The method of claim 1,
Wherein the probability distribution includes at least one of a normal distribution, a beta distribution, and a Poisson distribution. As a method for an advertisement providing apparatus to provide an advertisement to a user,
Managing exposure and performance histories of each of the plurality of advertisements,
Calculating a probability distribution of performance measures for each advertisement based on an exposure and a performance history of each of the plurality of advertisements,
Extracting a measurement index value of each advertisement based on a probability distribution of the advertisement-specific performance index, and
Selecting an advertisement to be provided to the user terminal based on the measurement index value of each advertisement
And an advertisement delivery method. 8. The method of claim 7,
Wherein the step of calculating the probability distribution of the performance index for each advertisement
An average and standard deviation associated with the performance indicators of each advertisement are calculated based on the exposure and performance histories related to the performance indicators of the respective advertisements and the probability distribution of the performance indicators is calculated for each advertisement based on the average and the standard deviation Delivery method. 8. The method of claim 7,
Wherein the probability distribution includes at least one of a normal distribution, a beta distribution, and a Poisson distribution. 8. The method of claim 7,
The performance indicator
A click rate per impression, and a response rate per impression. 8. The method of claim 7,
The step of extracting the measurement index value of each advertisement
And when the advertisement request is received from the user terminal, extracting a measurement index value of each advertisement based on a probability distribution of the advertisement measurement index. 12. The method of claim 11,
The step of extracting the measurement index value of each advertisement
The method according to claim 1 or 2, wherein when the first advertisement request is received, a first performance index value is extracted from a probability distribution of a performance index corresponding to the first advertisement, A method of providing an advertisement for extracting a measurement index value. 8. The method of claim 7,
Updating a probability distribution of a performance index corresponding to the arbitrary advertisement when the exposure and the performance history of the arbitrary advertisement are updated among the plurality of advertisements
Further comprising the steps of:

Images