KR102194031B1 - Technique for reducing advertising fraud - Google Patents

Abstract

A computer program stored in a computer-readable storage medium, the computer program comprising instructions for causing a server to perform the following operations, the operations being: at least one in at least one action block recorded in a first blockchain network Recognizing the user's action log; Calculating an advertisement performance value of a first publisher among a plurality of publishers using the action log; By transmitting the advertisement performance value of the first publisher to at least one action node included in the first blockchain network, a plurality of action nodes verify the validity of the advertisement performance value through a consensus algorithm, and the advertisement performance Causing the plurality of action nodes to record the advertisement performance values in a transaction in an action block as the validity of the values is verified; Recognizing a legality value of the first publisher based on the advertisement performance value of the first publisher; And when the legality value of the first publisher exceeds a preset threshold, restricting access to the advertising network of the first publisher. A computer program stored on a computer-readable storage medium comprising a.

Description

Techniques to reduce advertising fraud {TECHNIQUE FOR REDUCING ADVERTISING FRAUD}

The present disclosure relates to a method of reducing advertising fraud using a block chain.

Recently, the online advertising market is growing rapidly. As the public's interest and use of online content increases, advertisers' needs for online advertising are also increasing. In particular, the online advertising market is evolving more and more rapidly through the development of convenient transaction methods and data-based performance measurement technologies through online.

Meanwhile, along with the growth of the online advertising market, the issue of advertising fraud has been cited as a hot issue. Ad fraud refers to the fact that a publisher who executes an advertisement by receiving a request from an advertiser acts negatively, such as generating fraudulent traffic, thereby reducing the effectiveness of the advertisement. Methods of advertising fraud may include Farms, Click Fraud, and Hiding Ads. Here, Farms is the most basic form of advertising fraud, and refers to a fraudulent technique that employs many people to install content through clicks or various accounts. In addition, Click Fraud is a method of exploiting the unit price measurement method (CPC) according to the number of clicks, and refers to a fraud method such as increasing the number of clicks by using programs such as macros or bots. In addition, Hiding Ads refers to a fraudulent technique that causes advertisers to pay for inefficient advertisements by displaying advertisements, but hiding them from the eyes of users.

The above-described examples are only a few of the numerous advertisement fraud methods, and a problem of consuming enormous cost and time due to such advertisement fraud arises from the standpoint of a company that wants to request an advertisement.

Therefore, there is a demand for a method of ensuring the transparency of online advertising execution and reducing advertising fraud.

Republic of Korea Patent Publication 10-2018-0104586 Republic of Korea Patent Publication 10-2018-0046930 US published patent US 2010-0082400

The present disclosure is conceived in response to the above-described background technology, and intends to provide a method of reducing advertisement fraud by enabling transparent and reliable advertisement execution using blockchain technology.

The technical problems of the present disclosure are not limited to the technical problems mentioned above, and other technical problems that are not mentioned will be clearly understood by those skilled in the art from the following description.

According to an embodiment of the present disclosure for solving the above-described problem, a computer program stored in a computer-readable storage medium, wherein the computer program includes instructions for causing a server to perform the following operations, and the operation They include: recognizing an action log of at least one user from at least one action block recorded in the first blockchain network; Calculating an advertisement performance value of a first publisher among a plurality of publishers using the action log; By transmitting the advertisement performance value of the first publisher to at least one action node included in the first blockchain network, a plurality of action nodes verify the validity of the advertisement performance value through a consensus algorithm, and the advertisement performance Causing the plurality of action nodes to record the advertisement performance values in a transaction in an action block as the validity of the values is verified; Recognizing a legality value of the first publisher based on the advertisement performance value of the first publisher; And when the legality value of the first publisher exceeds a preset threshold, restricting access to the advertising network of the first publisher. Includes.

In addition, the operation of recognizing an action log of at least one user in the at least one action block recorded in the first blockchain network may include identifying each of the at least one user as an intrinsic user and a non-intrinsic user by using the action log. The operation separated by; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log includes the number of genuine users, the revenue according to the genuine users, the number of non-genuine users, and the non-genuine users. Calculating the return on advertising spend (ROAS) value by using the corresponding profit and the first weight value set for the genuine user; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log includes the number of genuine users, the revenue according to the genuine users, the number of non-genuine users, and the non-genuine users. Calculating a return on advertising spend (ROAS) value by using the corresponding profit and a first weight value set for the genuine user; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log includes the number of genuine users, the revenue according to the genuine users, the number of non-genuine users, and the non-genuine users. Calculating an ROAS value by using the corresponding profit, a first weight value set for the intrinsic user, and a second weight value set for the non-intrinsic user; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log includes the number of clicks of the advertisement of the genuine user, the number of clicks of the advertisement of the non-genuine user, and a third weight set for the genuine user. Calculating a conversion rate using the value; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log includes the number of clicks of the advertisement of the genuine user, the number of clicks of the advertisement of the non-genuine user, and a third weight set for the genuine user. Calculating a conversion rate using a value and a fourth weight value set for the non-intrinsic user; Includes.

In addition, the operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log is based on the first information on the number of suspicious reports of the first publisher included in the action log. The operation of calculating; Including, the first information includes at least one of the number of suspicious reports on the number of advertisements by the first publisher, the number of suspicious reports on the advertisement content, and the number of suspicious reports on the advertisement period.

In addition, the action log of the at least one user may include user identification information for each of the at least one user, time information at which the event occurred, information on the terminal where the event was executed, content payment information related to the event, and The operation of dividing each of the at least one user into a genuine user and a non-authentic user using at least one of the usage of the content related to the event and using the action log includes: the user identification information, the event occurrence Using at least one of the time information, information on the terminal on which the event was executed, the content payment information related to the event, and the usage amount of the content related to the event, each of the at least one user is identified as a genuine user and a non-genuine user. The operation of classifying by user; Includes.

In addition, the event includes at least one of an advertisement click event, an application installation event, and a preset action event.

In addition, the operation of transmitting the action log to at least one action node included in the first blockchain network may include generating a first transaction signed with a private key of an advertisement management server; And transmitting the generated first transaction to the at least one action node. Further comprising: a first transaction input value including identification information generated from the private key; And a first transaction output value including address information in the first blockchain network in which the action log and the first transaction including the action log are to be stored. Includes.

In addition, the operation of recognizing the legality value of the first publisher based on the advertisement performance value of the first publisher includes at least one of an advertisement fraud value, an ROAS value, and a conversion rate included in the advertisement performance value. Calculating the legality value by calculating using Includes.

In addition, the legitimacy numerical calculation model inputs a training data set including at least one of a training advertisement fraud value, a training ROAS value, and a training conversion rate into each of one or more input nodes included in an input layer of one or more network functions, and the The output of the legitimacy numerical calculation model is derived by propagating from the input layer to the output layer through one or more hidden layers through a link connected to each input node, and the learning legitimacy value and the legitimacy value labeled in each of the training data sets By comparing the output of the calculation model to derive an error, and propagating the derived error from the output layer of one or more network functions of the legality numerical calculation model to the input layer through the one or more hidden layers, to each link. It is a model created by updating the set weights and learning the one or more network functions.

In addition, collecting a new action log for each of at least one user at a preset time interval; And when the new action log is verified through a consensus algorithm by transmitting the new action log to the at least one action node included in the first blockchain network, a new action block is generated in each of the plurality of action nodes. Causing the new action log to be recorded in the new action block; Includes.

In addition, the operation of transmitting the new action log to at least one action node included in the first blockchain network may include generating a second transaction signed with a private key of an advertisement management server; And transmitting the generated second transaction to the at least one action node. And the second transaction includes: a second transaction input value including identification information generated from the private key; And a second transaction output value including address information in the first blockchain network in which the second transaction including the new action log and the new action log will be stored. Includes.

In addition, when the legality value of the first publisher is recognized, the legality value is transmitted to at least one legality numerical record node included in the second blockchain network, so that a plurality of legality numerical record nodes can be used through a consensus algorithm. Verifying the validity of the legality value, and causing the plurality of legality number recording nodes to record the legality value as a transaction in the legality number recording block as the validity of the legality value is verified; Includes.

In addition, when the legality value of the first publisher is recognized, the operation of transmitting the legality value to at least one legality value recording node included in the second blockchain network includes a private key of the advertisement management server. Generating a third transaction signed with ); And transmitting the generated third transaction to the at least one action node. And the third transaction includes: a third transaction input value including identification information generated from the private key; A third transaction output value including address information in the second blockchain network in which the legality value and the third transaction including the legality value are to be stored; Includes.

In addition, the operation of recognizing the legality value of the first publisher based on the advertisement performance value of the first publisher may include, when the first publisher requests access to the advertisement network, the legality value of the first publisher The operation of recognizing; Includes.

In addition, when the legality value of the first publisher exceeds a preset threshold, the operation of restricting access to the advertising network of the first publisher may include the legality value of the first publisher Limiting the access of the first publisher to the advertisement network for a preset time from when it is recognized that it exceeds a set threshold value; And when the preset time has elapsed, canceling the restriction on the access of the first publisher to the advertisement network. Includes.

In addition, analyzing the action log to obtain departure information of the at least one user; Generating a departure prediction model using the departure information; And predicting a user churn rate and a departure time point of each of the at least one user by calculating the action log using the departure prediction model. Includes.

In addition, the operation of generating a departure prediction model using the departure information includes inputting the action log as training data to each of one or more input nodes included in the input layer of one or more network functions, and connected to each of the input nodes. Deriving an output of the departure prediction model by propagating from the input layer to an output layer through one or more hidden layers through a link; And comparing the data labeled in the training data with the output of the departure prediction model to derive an error, and the derived error from the output layer of one or more network functions of the departure prediction model through the one or more hidden layers. Learning the one or more network functions by updating weights set for each link by propagating to an input layer; Includes.

In addition, the block header of each of the at least one action block includes a first Merkle root value for information stored in a transaction of each of the at least one action block.

Further, the block header of the legality number recording block includes a second Merkle root value for the legality value of the first publisher and the legality value of at least one other publisher other than the first publisher.

Further, a computer program stored in a computer-readable storage medium, wherein the computer program includes instructions for causing a server to perform the following operations, the operations being: at least one in at least one action block recorded in the blockchain network. Recognizing the user's action log; Calculating an advertisement performance value of a first publisher among a plurality of publishers using the action log; By transmitting the advertisement performance value of the first publisher to at least one action node included in the blockchain network, a plurality of action nodes verify the validity of the advertisement performance value through a consensus algorithm, and the advertisement performance value Causing the plurality of action nodes to record the advertisement performance figures in a transaction in a first action block as validity is verified; Recognizing a legality value of the first publisher based on the advertisement performance value of the first publisher; When the legality value of the first publisher is generated, by transmitting the legality value to at least one legality value recording node included in the blockchain network, a plurality of legality value recording nodes can determine the legality value through a consensus algorithm. Verifying validity, and causing the plurality of legality number recording nodes to record the legality value as a transaction in a legality number recording block as the validity of the legality value is verified; And when the legality value of the first publisher exceeds a preset threshold, restricting access to the advertising network of the first publisher. Includes.

Technical solutions obtainable in the present disclosure are not limited to the above-mentioned solutions, and other solutions that are not mentioned are clearly to those of ordinary skill in the technical field to which the present disclosure belongs from the following description. It will be understandable.

The present disclosure can reduce advertisement fraud by enabling transparent and reliable advertisement execution.

The effects obtainable in the present disclosure are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those of ordinary skill in the art from the following description. .

Various aspects are now described with reference to the drawings, wherein like reference numbers are used collectively to refer to like elements. In the examples that follow, for illustrative purposes, a number of specific details are presented to provide a holistic understanding of one or more aspects. However, it will be apparent that such aspect(s) may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form to facilitate describing one or more aspects.
1 depicts an exemplary ad fraud reduction system for performing a method of reducing ad fraud according to an embodiment of the present disclosure.
2 is a flowchart illustrating an example of a method of restricting access to an ad network of a specific publisher according to some embodiments of the present disclosure.
3 is a flowchart illustrating an example of a method of recording a new action log for each user in a blockchain network according to some embodiments of the present disclosure.
4 is a flow chart for explaining an example of a method of recognizing a legality value of a first publisher according to some embodiments of the present disclosure.
5 is a flowchart illustrating an example of a method of releasing access restrictions after restricting access to an advertisement network of a first publisher according to some embodiments of the present disclosure.
6 is a flowchart illustrating an example of a method of analyzing at least one user's departure by analyzing an action log according to some embodiments of the present disclosure.
7 is a schematic diagram illustrating a network function according to some embodiments of the present disclosure.
8 is a diagram for describing a legality numerical calculation model and a departure prediction model according to some embodiments of the present disclosure.
9 exemplarily shows a connection structure of blocks stored in a first blockchain network according to some embodiments of the present disclosure.
10 is an exemplary diagram illustrating an example of a method of generating a first Merkle root value according to some embodiments of the present invention.
11 exemplarily shows a connection structure of blocks stored in a second blockchain network according to some embodiments of the present disclosure.
12 is an exemplary diagram illustrating an example of a method of generating a second Merkle root value according to some embodiments of the present invention.
13 shows a general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure may be implemented.

Various embodiments and/or aspects are now disclosed with reference to the drawings. In the following description, for illustrative purposes, a number of specific details are disclosed to aid in an overall understanding of one or more aspects. However, it will also be appreciated by those of ordinary skill in the art that this aspect(s) may be practiced without these specific details. The following description and the annexed drawings set forth in detail certain illustrative aspects of one or more aspects. However, these aspects are exemplary and some of the various methods in the principles of the various aspects may be used, and the descriptions described are intended to include all such aspects and their equivalents. Specifically, as used herein, "an embodiment", "example", "aspect", "example" and the like are not construed as having any aspect or design being better or advantageous than other aspects or designs. May not.

Hereinafter, the same or similar components are assigned the same reference numerals regardless of the reference numerals, and redundant descriptions thereof are omitted. In addition, in describing the embodiments disclosed in the present specification, when it is determined that detailed descriptions of related known technologies may obscure the subject matter of the embodiments disclosed in the present specification, detailed descriptions thereof will be omitted. In addition, the accompanying drawings are for easy understanding of the embodiments disclosed in the present specification, and the technical spirit disclosed in the present specification is not limited by the accompanying drawings.

Although the first, second, and the like are used to describe various devices or components, it goes without saying that these devices or components are not limited by these terms. These terms are only used to distinguish one device or component from another device or component. Therefore, it goes without saying that the first element or element mentioned below may be a second element or element within the technical idea of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used in the present specification may be used as meanings that can be commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not interpreted ideally or excessively unless explicitly defined specifically.

In addition, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or”. That is, unless specified otherwise or is not clear from the context, "X employs A or B" is intended to mean one of the natural inclusive substitutions. That is, X uses A; X uses B; Or, when X uses both A and B, “X uses A or B” can be applied to either of these cases. In addition, the term "and/or" as used herein should be understood to refer to and include all possible combinations of one or more of the listed related items.

In addition, the terms "comprising" and/or "comprising" mean that the corresponding feature and/or element is present, but excludes the presence or addition of one or more other features, elements and/or groups thereof. It should be understood as not. In addition, unless otherwise specified or when the context is not clear as indicating a singular form, the singular in the specification and claims should be interpreted as meaning "one or more" in general.

In addition, the terms “information” and “data” as used herein may often be used interchangeably with each other.

When a component is referred to as being "connected" or "connected" to another component, it is understood that it is directly connected to or may be 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.

The suffixes "module" and "unit" for constituent elements 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 a distinct meaning or role by themselves.

Objects and effects of the present disclosure, and technical configurations for achieving them will become apparent with reference to embodiments described in detail later with reference to the accompanying drawings. In describing the present disclosure, when it is determined that a detailed description of a well-known function or configuration may unnecessarily obscure the subject matter of the present disclosure, a detailed description thereof will be omitted. In addition, terms to be described later are terms defined in consideration of functions in the present disclosure and may vary according to the intention or custom of users or operators.

The scope of rights to the method in the claims of the present disclosure is generated by the functions and features described in each step, and unless a precedence relationship of the order is specified in each step constituting the method, the claim It is not affected by the order of description of each step in the range. For example, in the claims described in a method including steps A and B, even if step A is described before step B, the scope of the rights is not limited to that step A must precede step B.

However, the present disclosure is not limited to the embodiments disclosed below, and may be implemented in various different forms. The present embodiments are provided only to make the present disclosure complete, and to completely inform the scope of the disclosure to those of ordinary skill in the art to which the present disclosure belongs, and the present disclosure is only defined by the scope of the claims. . Therefore, the definition should be made based on the contents throughout this specification.

1 depicts an exemplary ad fraud reduction system for performing a method of reducing ad fraud according to an embodiment of the present disclosure.

Referring to FIG. 1, the advertisement fraud reduction system includes a user terminal 100, a first blockchain network 200, a second blockchain network 300, an advertisement network 400, a publisher 500, and an advertisement management server ( 600) and a network 700. However, since the above-described components are not essential to implement the advertising fraud reduction system, the advertising fraud reduction system may have more or fewer components than the above-listed components.

According to some embodiments of the present disclosure, a user terminal 100, a first blockchain network 200, a second blockchain network 300, an advertisement network 400, a publisher 500, and an advertisement management server 600 May mutually transmit and receive data for the advertisement fraud reduction system according to some embodiments of the present disclosure through the network 700.

The user terminal 100 may include a personal computer (PC), a notebook (note book), a mobile terminal, a smart phone, a tablet PC, etc. owned by the user. , It may include all types of terminals that can access wired/wireless networks.

The user terminal 100 may be a terminal used by a user who subscribes and clicks an advertisement issued by the publisher 500 through the network 700. In addition, advertisements in this disclosure include arbitrary types of advertisements such as Searching Ad, Display Ad, Mobile Ad, Viral Ad, and Direct Mailing Ad. Can include. However, the present disclosure is not limited thereto, and various types of advertisements may be advertisements described in some embodiments of the present disclosure.

The first blockchain network 200 includes a user's action log made through the user terminal 100 and a value calculated by the action log (e.g., advertising fraud value, ROAS (Return on advertising spend), and conversion rate). It can mean a network that records advertisement performance figures including rate), etc.) through a blockchain.

Here, the action log is information on an arbitrary type of action performed by a user through the user terminal 100 or an advertisement suspicious report performed by an advertiser through an advertiser terminal (not shown) (for example, reporting suspicious advertisement fraud. This may mean recording information about the advertiser who has done so, information about the publisher who is the target of suspicious advertisement fraud. For example, the action log shows the action of a user clicking an advertisement through the user terminal 100, downloading content, performing membership registration, performing installation, performing uninstall, and member withdrawal. When an action occurs, it may mean that information on the action has been recorded. A detailed description of the action log will be described later with reference to FIG. 2.

In addition, each of the action nodes included in the first blockchain network 200 may record an advertisement performance value including at least one of an advertisement fraud value, an ROAS value, and a conversion rate in the action block as a transaction through a consensus algorithm. . Here, the advertisement performance value may be a value that infers the degree of fraud based on the advertisement performance of each publisher.

The advertisement fraud value may be a value based on the number of suspicious reports of the first publisher included in the action log.

The ROAS value may be a value representing the advertising yield. For example, the ROAS value may be a value representing the advertising cost invested by the advertiser relative to sales generated by advertising.

The conversion rate may mean a percentage of users who take an action induced by an advertiser among those who click an advertisement. For example, if an advertiser induces a user to download content through an advertisement intervened by an advertiser, it may be a number indicating a user who downloaded the content compared to the total number of clicks on the advertisement.

In addition, an advertisement management server 600 may be included in a node included in the first blockchain network 200.

A detailed description of the above-described advertisement fraud value, ROAS value, and conversion rate will be described later in detail with reference to FIG. 2.

Meanwhile, an action block is a unit in which action logs and/or advertisement performance figures are stored in the first blockchain network 200, and may form a block chain by being connected to adjacent blocks.

When the action log and/or the advertisement performance figures are recorded in the first blockchain network 200, it is virtually impossible for the advertisement performance figures to be forged or altered. Accordingly, it is possible to prevent the malicious publisher 500 from performing advertisement fraud by hacking and modifying the action log and/or advertisement performance figures through the network 700.

Not only are the action logs used when calculating the advertising performance figures are recorded as action blocks by the plurality of action nodes of the first blockchain network 200, but when the advertising performance figures are calculated, the advertising performance figures are also the first blockchain. Since a plurality of action nodes of the network 200 are recorded as action blocks, if one is forged or altered, all information must be corrected, so forgery or alteration may be virtually impossible. Accordingly, the reliability of the advertising fraud reduction system can be improved.

The second blockchain network 300 may mean a network that records the legality value of at least one publisher 500 in a blockchain.

At this time, each of the legality numerical record nodes included in the second blockchain network 300 may record the legality numerical record of at least one publisher in a transactional legality numerical record block through a consensus algorithm. Here, the legality value recording block is a unit in which the legality value of at least one publisher 500 is stored in the second block chain network 300, and may form a block chain by being connected to adjacent blocks.

In addition, an advertisement management server 600 may be included in a node included in the second blockchain network 300.

In addition, when the legality value for at least one publisher 500 is recorded in the second blockchain network 300, it is virtually impossible to forge or alter the legality value for at least one publisher 500. Accordingly, it is possible to prevent the malicious publisher 500 from hacking and modifying the legality figures through the network 700.

Meanwhile, not only the plurality of action nodes of the first blockchain network 200 record at least one of the advertisement fraud value, ROAS value, and conversion rate used when calculating the legality value as an action block, but also the legality value as the second. Since a plurality of legitimacy numeric recording nodes of the blockchain network 300 are recording legitimacy numeric recording blocks, if one is forged or altered, all information must be modified, so forgery or alteration may be virtually impossible. Accordingly, the reliability of the advertising fraud reduction system can be improved.

According to some embodiments of the present disclosure, the first blockchain network 200 and the second blockchain network 300 are separately expressed as separate networks, but according to an embodiment of the present disclosure, the integrated blockchain network It may be operated as (300, 400).

According to some embodiments of the present disclosure, the first blockchain network 200 may be operated in a form existing inside the second blockchain network 300. In addition, the second blockchain network 300 may be operated in a form existing inside the first blockchain network 200. Hereinafter, the first block chain network 200 and the second block chain network 300 will be described separately, but even when the first block chain network 200 and the second block chain network 300 are integrated, in the present disclosure The described embodiments can be applied.

In addition, an advertisement management server 600 may be included in a node included in the blockchain networks 200 and 300.

On the other hand, in some embodiments of the present disclosure, a plurality of nodes (action nodes, legality numerical record nodes) included in the blockchain networks 200 and 300 may be a full block chain node, It can also be a Lightweight Node.

A full block chain node can contain all block information from the first block of the block chain to a block that is currently newly created. In addition, the full blockchain node collects and stores all blockchain information, and can verify the received block to add a new block.

The Lightweight Node does not have the originals of all block information, and may include only header information. In order for the Lightweight Node to verify the transaction, it can perform Simple Payment Verify (SPV).

For example, a lightweight node can request block information from a full blockchain node, and the authentication content of a transaction can be verified through Merkle Root.

However, for convenience of explanation, it is assumed that a plurality of content nodes included in the blockchain networks 200 and 300 are full blockchain nodes and will be described below.

The consensus algorithm performed in the blockchain networks 200 and 300 according to some embodiments of the present disclosure is a PoW (Proof of Work) algorithm, PoS (Proof of Stake) algorithm, DPoS (Delegated Proof of Stake) algorithm, PBFT (Practical Byzantine Fault Tolerance) algorithm, Delegated Byzantine Fault Tolerance (DBFT) algorithm, Redundant Byzantine Fault Tolerance (RBFT) algorithm, Sieve algorithm, Tendermint algorithm, Paxos algorithm, Raft algorithm, PoA (Proof of Authority) algorithm and/or PoET (Proof of Elapsed Time) algorithm may be included. However, the present invention is not limited thereto, and an algorithm different from the above-described algorithm may be used as the consensus algorithm.

In addition, by using the consensus algorithm of blockchain technology, action logs, advertisement fraud values, ROAS values, conversion rates and legality values, etc., are recorded in blocks to prevent forgery or falsification of information related to advertisement fraud by the publisher 500. There is an effect.

Meanwhile, a block chain network structure according to some embodiments of the present disclosure may be a public type or a private type. In the case of public blockchain network structure, there is a disadvantage that it is inefficient in terms of consensus time, transaction processing speed, and efficiency of use of computing resources, since it is necessary to perform verification work on all nodes to verify transactions. Because it is possible, there may be advantages in transparency and integrity of the stored data. In addition, in the case of a private (or consortium) block chain, since the operating entity is clear, there is an advantage of being efficient in terms of consensus time, transaction processing speed, and efficiency of use of computing resources, but transparency about the transaction processing process and results. There may be advantages in terms of

The consensus algorithm in some embodiments of the present disclosure may be a private or consortium-type blockchain network, but is not limited thereto, in order to improve the user experience through efficient use of computing resources and increase in transaction processing speed. Depending on the implementation mode, public type may be used as the consensus algorithm in order to further emphasize the transparency of the transaction.

According to some embodiments, nodes in the blockchain networks 200 and 300 may be operated by a blockchain core package according to a hierarchical structure. The above structure defines the structure of data handled in the blockchain networks 200 and 300, a data layer that manages the data, performs mining to verify the validity of the block and creates a block, and is paid to the miners during the mining process. A consensus layer in charge of processing fees, a P2P network protocol, a hash function, a common layer implementing and managing digital signatures, encoding and common storage, and an application layer in which various applications are created, processed, and managed may be included.

The advertisement network 400 is a network that connects the advertiser and the publisher 500, and may mean a platform through which a plurality of advertisers can execute advertisements to various publishers 500 at once. That is, the advertiser may use the advertisement network 400 to search for a publisher 500 to perform an advertisement, and the publisher 500 may use the advertisement network 400 to find an advertiser.

The publisher 500 may refer to a program, platform, website, and company that collects information or content online according to a predetermined schedule and performs a service that provides information to a user.

According to some embodiments of the present disclosure, publisher 500 may include web publishers, application publishers, mobile publishers, and the like. However, the present invention is not limited thereto, and various platforms and programs may be used as the publisher 500.

The advertisement management server 600 restricts access to the advertisement network 400 of a specific publisher 500 or terminates the advertisement function of a specific publisher 500 by using the legality value recorded in the second blockchain network 300. can do. Hereinafter, a detailed description will be provided with reference to FIGS. 2 to 6.

The advertisement management server 600 may include any type of computer system or computer device, such as, for example, a microprocessor, mainframe computer, digital processor, portable device and device controller, and the like.

The advertisement management server 600 may transmit a query or a transaction to at least one of the first blockchain network 200 and the second blockchain network 300. The query in the present disclosure may be used to inquire an action log or advertisement performance value stored in the first blockchain network 200. In addition, the query in the present disclosure may be used to query the legality value stored in the second blockchain network 300. In addition, the transaction in the present disclosure may be used to perform an update (modification/change/delete/add) on data recorded in the first blockchain network 200 and the second blockchain network 300.

The advertisement management server 600 may include a processor 610, a storage unit 620 and a communication unit 630. However, the above-described components are not essential for implementing the advertisement management server 600, and thus the advertisement management server 600 may have more or fewer components than the components listed above. Here, each component may be composed of a separate chip, module or device, or may be included in a single device.

The processor 610 may typically process the overall operation of the advertisement management server 600. The processor 610 provides or processes appropriate information or functions to a user by processing signals, data, information, etc. that are input or output through the above-described components or by running an application program stored in a memory (not shown in FIG. 1). can do.

In addition, when the processor 610 recognizes at least one user action, the processor 610 may generate an action log and transmit it to at least one action node included in the first blockchain network 200. In this case, when the action log is verified through the consensus algorithm, an action block may be generated in each of a plurality of action nodes, and the action log may be recorded in the action block. Meanwhile, the processor 610 generates a transaction signed with the private key of the advertisement management server 600 when transmitting the action log to at least one action node, and transmits the generated transaction to the at least one action node. ) Can be controlled. Here, the identification information generated from the private key of at least one user may be recorded in the input value of the transaction. In addition, address information in the first blockchain network 200 in which a transaction including an action log and an action log of at least one user is to be stored may be recorded in the output value of the transaction.

The storage unit 620 may store information in an arbitrary form generated or determined by the processor 610 and information in an arbitrary form received by the communication unit 630 in the advertisement management server 600. The storage unit 620 is a device that stores data on a magnetic disk such as a hard disk drive (HDD, Hard Disk Drive), a solid state drive (SSD, Solid State Drive), a semiconductor memory such as a memory card, especially a nonvolatile memory. It may include a device for storing data.

The communication unit 630 is between the advertisement management server 600 and the communication system, between the advertisement management server 600 and the user terminal 100, between the advertisement management server 600 and the first blockchain network 200, an advertisement management server One or more modules that enable communication between 600 and the second blockchain network 300, between the advertisement management server 600 and the advertisement network 400, or between the advertisement management server 600 and the publisher 500. Can include.

The communication unit 630 may include at least one of a mobile communication module, a wired Internet module, and a wireless Internet module.

The network 700 according to the embodiments of the present disclosure includes a public switched telephone network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), and Very Various wired communication systems such as High Speed DSL), Universal Asymmetric DSL (UADSL), High Bit Rate DSL (HDSL), and local area network (LAN) can be used.

In addition, the network 700 presented here is CDMA (Code Division Multi Access), TDMA (Time Division Multi Access), FDMA (Frequency Division Multi Access), OFDMA (Orthogonal Frequency Division Multi Access), SC-FDMA (Single Carrier- FDMA) and other systems can be used.

The network 700 according to the embodiments of the present disclosure may be configured regardless of its communication mode, such as wired or wireless, and various types of short-range communication networks (PAN: Personal Area Network), local area networks (WAN: Wide Area Network), etc. It can be configured as a communication network. In addition, the network may be a known World Wide Web (WWW), and may use a wireless transmission technology used for short-range communication such as infrared (IrDA: Infrared Data Association) or Bluetooth (Bluetooth).

The techniques described herein may be used not only in the networks mentioned above, but also in other networks.

2 is a flowchart illustrating an example of a method of restricting access to an ad network of a specific publisher according to some embodiments of the present disclosure.

Referring to FIG. 2, the processor 610 of the advertisement management server 600 may recognize an action log of at least one user from at least one block recorded in the first blockchain network 200 (S110). Here, the action log includes at least one of user identification information for each of at least one user, information on the time when the event occurred, information on the terminal where the event was triggered, information on payment of content related to the event, and usage of content related to the event. can do.

Specifically, the processor 610 transmits an action log request query to at least one action node included in the first blockchain network 200 in order to recognize at least one block included in the first blockchain network 200. I can. Further, the processor 610 may recognize the action log by receiving the action log of at least one user recorded in the block from the at least one action node. However, the method of recognizing the action log is not limited thereto.

According to some embodiments of the present disclosure, in step S110, the processor 610 may classify each of at least one user into a genuine user and a non-genuine user by using an action log. Here, the true user executes the action expected by the advertiser according to necessity regardless of the price (e.g., payback, accumulation of points) that can be obtained from clicking an advertisement, downloading the content through the advertisement, and executing the content. Can mean user. In addition, a user who is not classified as a genuine user described above may be classified as a non-authentic user. However, it is not limited thereto.

Specifically, the processor 610 includes at least one of user identification information for each of at least one user, information about the time when the event occurred, information on the terminal where the event was executed, content payment information related to the event, and content usage related to the event. Each of the at least one user may be classified into a genuine user and a non-genuine user by using. Here, the event is at least one of an advertisement click, an application installation, an additional data download, a guest login, an account login, an application uninstall, and a preset action (for example, an action indicating that the level of a character in the game reaches a specific level after installing the game). It may include. However, the event is not limited thereto.

Meanwhile, the processor 610 may calculate the advertisement performance value of the first publisher among the plurality of publishers 500 by using the recognized action log of at least one user (S120). Here, the advertisement performance value may be a value including at least one of an advertisement fraud value, an ROAS value, and a conversion rate. Therefore, the processor 610 calculates the advertisement performance value by calculating at least one of the advertisement fraud value, ROAS value, and conversion rate of the first publisher among the plurality of publishers 500 using the recognized action log of at least one user. I can.

The processor 610 may calculate the advertisement fraud value of the first publisher by using the action log recognized in step S110.

Specifically, the processor 610 may calculate the advertisement fraud value based on the first information on the number of suspicious reports of the first publisher included in the action log recognized in step S110. Here, the first information may include the number of suspicious reports on the number of advertisements by the first publisher, the number of suspicious reports on the advertisement content, the number of suspicious reports on the advertisement period, and the number of suspicious reports on the advertisement location. .

For example, if the processor 610 receives the report of suspicion about the number of advertisements of the first publisher through the communication unit 630 three times, the storage unit 620 to store the advertisement fraud value for the first publisher as -3. ) Can be controlled. In addition, the processor 610 may control the storage unit 620 to store the advertisement fraud value for the first publisher as -5, if a suspicious report on the advertisement content of the first publisher is received two more times. .

According to some embodiments of the present disclosure, the subject of reporting suspicion to the first publisher may include an advertiser, a second publisher, and at least one user using the advertisement network 400.

According to some embodiments of the present disclosure, the processor 610 may calculate the ROAS value of the first publisher by using the action log recognized in step S110.

Specifically, the processor 610 may recognize the number of users who clicked the advertisement, and calculate the ROAS value through revenue and advertisement cost according to each user who clicked the recognized advertisement.

An example of generating the ROAS value of the first publisher based on the number of users who clicked the advertisement and the revenue according to each user who clicked the advertisement is expressed as Equation 1 below.

Here, ROAS may be an advertisement return rate, P may be the number of users who click the advertisement, R may be revenue according to each user who clicked the advertisement, and A _C may be an advertisement cost.

According to some embodiments of the present disclosure, when the processor 610 distinguishes between genuine users and non-genuine users in step S110, the number of genuine users, revenue according to genuine users, number of non-genuine users, non-genuine users The ROAS value may be calculated by using the first weight value set for the profit according to and the genuine user.

An example of calculating the ROAS value using the first weight value set for the true user is expressed as Equation 2 below.

Here, ROAS ₁ is the ROAS value weighted by the first weight value, α is the first weight value, P ₁ is the number of genuine users, R ₁ is the revenue according to the genuine user, P ₂ is the number of non-genuine users, R ₂ is Revenue according to non-genuine users, A _C may be advertising expenses.

According to Equation 2, when calculating the ROAS value, since only the profit according to the genuine user is weighted, the accuracy of calculating the ROAS value may be increased.

In addition, according to some embodiments of the present disclosure, if the processor 610 distinguishes between genuine users and non-genuine users in step S110, the number of genuine users, revenue according to genuine users, number of non-genuine users, and ratio The ROAS value may be calculated using the revenue according to the genuine user, the first weight value set for the genuine user, and the second weight value set for the non-genuine user. Here, the size of the first weight value may be larger than the size of the second weight value. Therefore, when calculating the ROAS value, since more weight is given to the profits according to genuine users, the accuracy of calculating the ROAS value can be improved.

An example of calculating the ROAS value using the first weight value set for the intrinsic user and the second weight value set for the non-intrinsic user is expressed as Equation 3 below.

는 제 2 웨이트 값, P1은 진성 유저의 수, R1은 진성 유저에 따른 수익, P2는 비진성 유저의 수, R₂는 비진성 유저에 따른 수익, A_C는 광고비 일 수 있다.Here, ROAS ₃ is the ROAS value weighted by the first weight value and the second weight value, α is the first weight value,

Is the second weight value, P1 is the number of genuine users, R1 is the revenue according to the genuine user, P2 is the number of non-genuine users, R ₂ is the revenue according to the non-genuine user, and A _C is the advertising cost.

According to some embodiments of the present disclosure, the processor 610 may calculate the conversion rate of the first publisher by using the action log recognized in step S110.

Specifically, the processor 610 may calculate a conversion rate based on the number of times the advertisement is clicked during the period in which the advertisement is opened through the first publisher and the number of users who take a preset action after clicking the advertisement.

An example of generating a conversion rate based on the number of clicks of an advertisement and the number of users who took a preset action is expressed as Equation 4 below.

Here, V _C may be the number of times the advertisement was clicked, and U may be the number of users who took a preset action.

According to some other embodiments of the present disclosure, in the case of discriminating between a genuine user and a non-genuine user in step S110, the processor 610 includes the number of clicks of an advertisement by a genuine user, the number of clicks of an advertisement by a non-genuine user, and set to the genuine user. The conversion rate can be calculated using the third weight value.

An example of generating the second conversion rate using the third weight value set for the genuine user is expressed as Equation 5 below.

Here, α may be a third weight value, U ₁ may be the number of clicks of an advertisement by an authentic user, U ₂ may be the number of clicks on an advertisement by a non-authentic user, and A _C may be an advertisement cost.

According to some other embodiments of the present disclosure, in the case of discriminating between a genuine user and a non-genuine user in step S110, the processor 610 includes the number of clicks of an advertisement by a genuine user, the number of clicks of an advertisement by a non-genuine user, The conversion rate may be calculated using the third weight value and the fourth weight value set for the non-intrinsic user.

Specifically, the processor 610 may calculate a third conversion rate by using the number of genuine users, the number of clicks of advertisements by genuine users, the number of non-genuine users, the number of clicks on advertisements by non-genuine users, and the fourth weight value. Here, the size of the third weight value may be larger than the size of the fourth weight value. Therefore, when calculating the ROAS value, since the number of clicks according to the genuine user is more weighted, the accuracy of calculating the conversion rate can be improved.

An example of generating a third conversion rate using the first weight value and the second weight value set for the genuine user is expressed as Equation 6 below.

는 제 4 웨이트 값, U₁은 진성 유저의 광고 클릭 횟수, U₂는 비진성 유저의 광고 클릭 횟수, A_C는 광고비 일 수 있다.Where α is the third weight value,

Is a fourth weight value, U ₁ is the number of clicks of an advertisement by an authentic user, U ₂ is the number of clicks on an advertisement by a non-authentic user, and A _C may be an advertisement cost.

The processor 610 may control the communication unit 630 to transmit the calculated advertisement performance value of the first publisher to at least one action node included in the first blockchain network (S130). When the advertisement management server 600 transmits the advertisement performance value to at least one action node, the plurality of action nodes verify the validity of the advertisement performance value through a consensus algorithm, and as the validity of the advertisement performance value is verified, the plurality of The action node of can record the advertisement performance in the action block as a transaction.

Specifically, each of the action nodes included in the first blockchain network 200 may record action information as a transaction in the action block through a consensus algorithm. In addition, action blocks may be connected by chains referred to in a plurality of blockchain technologies. In addition, when the advertisement performance figures are recorded in the first blockchain network 200, it is virtually impossible for the advertisement performance figures to be forged or altered. Accordingly, it is possible to prevent malicious publishers from hacking and modifying advertisement performance figures, and information contained in recorded transactions can be kept in a state where integrity can be guaranteed.

The processor 610 may recognize the legality value of the first publisher based on the advertisement performance value of the first publisher transmitted to the action node (S140).

For example, the processor 610 may recognize the legality value of the first publisher when there is a request for access to the advertisement network 400 of the first publisher.

As another example, the processor 610 may recognize the legality value of the first publisher at a preset time interval.

According to some embodiments, an algorithm for calculating a publisher's legality value based on at least one of an advertisement fraud value, an ROAS value, and a conversion rate may be stored in the memory of the advertisement management server 600. The processor 610 may calculate a publisher's legality value by calculating at least one of an advertisement fraud value, an ROAS value, and a conversion rate using the above-described algorithm.

According to some other embodiments, the processor 610 may calculate a legality value by calculating at least one of an advertisement fraud value, an ROAS value, and a conversion rate using a legality value calculation model. Here, the legitimacy numerical calculation model may be a model trained by using a training data set including at least one of a training advertisement fraud value, a training ROAS value, and a training conversion rate.

Specifically, when training the model for calculating the legality numerical value, a training data set including at least one of a training ad fraud value, a training ROAS value, and a training conversion rate is input to each of one or more input nodes included in the input layer of one or more network functions. In addition, it is possible to derive the output of the legality numerical calculation model by propagating from the input layer to the output layer through one or more hidden layers through a link connected to each input node. And, the error derived by comparing the output of the model for calculating the legality value and the learning legality value labeled in each of the training data sets is propagated from the output layer of one or more network functions of the legality value calculation model to the input layer through at least one hidden layer. The weight set for each link may be updated. As the weights are updated in this way, one or more network functions may be learned, and the generated model may be a legitimacy numerical calculation model. Hereinafter, a detailed description of the legality numerical calculation model will be described later in more detail with reference to FIGS. 7 to 8.

However, the present invention is not limited thereto, and the model for calculating the legality numerical value may be a model trained through semi-teacher learning or comparative history learning.

Meanwhile, when the legality value is recognized in step S140, the processor 610 transmits the legality value to at least one legality number recording node included in the second blockchain network 300, so that the plurality of legality number recording nodes The validity of the legality figure is verified through the consensus algorithm, and as the validity of the legality figure is verified, it can cause multiple legality figure recording nodes to record the legality figure in a transaction in the legality figure record block.

On the other hand, when the processor 610 transmits the legality value to at least one legality value recording node, it generates a third transaction signed with the private key of the advertisement management server 600, and the generated third transaction is at least one legality value. The communication unit 630 may be controlled to transmit to the recording node. Here, the identification information generated from the private key of the advertisement management server 600 may be recorded in the input value of the third transaction. In addition, address information in the second blockchain network 300 in which the third transaction including the legality value and the legality value of the publisher 500 of at least one user will be stored may be recorded in the output value of the third transaction. have.

Meanwhile, the processor 610 may recognize whether or not the legality value calculated for the first publisher exceeds a preset threshold (S150). Here, the preset threshold may be a value set by an administrator.

When the processor 610 recognizes that the legality value generated by using the advertisement fraud value of the first publisher recorded in the second blockchain network 300 exceeds a preset threshold (S150, Yes), the first publisher It is possible to restrict access to the ad network 400 (S160).

As an example, when the processor 610 recognizes that the calculated legality value for the first publisher exceeds a preset threshold, the processor 610 may perform the first time during a preset time period (eg, 5 hours, 5 days, 1 year, etc.). It is possible to limit the publisher's access to the ad network 400. Further, the processor 610 may cancel the restriction on access to the advertisement network of the first publisher when a preset time has elapsed.

As another example, when the processor 610 recognizes that the calculated legality value for the first publisher exceeds a preset threshold value, the processor 610 sets an access restriction period based on the size of the legality value, and is restricted during the access restriction period. 1 It is possible to limit the access of the publisher to the ad network 400.

Meanwhile, when the processor 610 recognizes that the calculated legality value for the first publisher is less than a preset threshold (S150, No), the first publisher may not restrict access to the advertisement network 400. That is, the processor 610 may allow the first publisher to access the advertisement network 400.

3 is a flowchart illustrating an example of a method of recording a new action log for each user in a blockchain network according to some embodiments of the present disclosure.

Referring to FIG. 3, the processor 610 may collect a new action log for each user at a preset time interval (eg, one hour interval) (S210). However, the operation of collecting new action logs for each user is not limited to the above-described example.

In addition, the processor 610 may control the communication unit 630 to transmit the collected new action log to at least one action node (S220).

Specifically, the processor 610 transmits the collected new action log to at least one action node included in the first blockchain network 200, so that the new action log is verified through a consensus algorithm, and a plurality of verified actions. A new action block can be created in each node to cause a new action log to be recorded in a new action block.

More specifically, the processor 610 generates a second transaction signed with the private key of the advertisement management server 600 in transmitting a new action log to at least one action node included in the first blockchain network 200. , It is possible to control the communication unit 630 to transmit the generated second transaction to at least one action node. Here, the identification information generated from the private key of at least one user may be recorded in the input value of the second transaction. In addition, address information in the first blockchain network 200 in which a second transaction including at least one user's new action log and the new action log will be stored may be recorded in the output value of the second transaction.

When a new action log is recorded in the action block of the first blockchain network 200, when calculating the ad fraud value, ROAS (Return on advertising spend) value, and conversion rate of a specific publisher in the future new action log Can be used.

As described above, when a new action log is collected, the processor 610 records a new action log in the first blockchain network 200 to prevent malicious publishers from hacking and modifying the new action log. , It can be kept in a state where the integrity of the recorded new action log can be guaranteed.

4 is a flow chart for explaining an example of a method of recognizing a legality value of a first publisher according to some embodiments of the present disclosure.

According to some embodiments of the present disclosure, the processor 610 may recognize the legality value of the first publisher in step S140 of FIG. 2 at a preset time interval.

Meanwhile, according to some other embodiments of the present disclosure, the processor 610 may recognize the legality value of the first publisher when receiving a request for access to the advertisement network 700 of the first publisher.

Specifically, referring to FIG. 4, the processor 610 transmits the advertisement performance value of the first publisher to the action node in step S130 of FIG. 2, and then requests access to the advertisement network 400 of the first publisher. Whether or not there is (S310).

When the processor 610 recognizes that there is a request for access to the advertisement network 400 from the first publisher, the processor 610 is recorded in the first blockchain network 200 whenever the first publisher wants to access the advertisement network 400. The first publisher's advertising performance figures can be used to recognize the first publisher's legality figures.

Whenever the processor 610 calculates at least one of the advertisement fraud value, ROAS value, and conversion rate included in the advertisement performance value of the first publisher, the process in the advertisement management server 600 As the throughput increases, the processing speed of the advertisement management server 600 may be slowed.

However, only when the processor 610 recognizes that there is a request for access to the ad network 400 of the first publisher, among the ad fraud values, ROAS value, and conversion rate of the first publisher recorded in the first blockchain network 200 If at least one is used to recognize the legality value of the first publisher, the process throughput can be reduced to some extent, and the process throughput in the advertisement management server 600 can be reduced, thereby increasing the process processing speed of the advertisement management server 600. Can be improved.

5 is a flowchart illustrating an example of a method of releasing access restrictions after restricting access to an advertisement network of a first publisher according to some embodiments of the present disclosure.

Referring to FIG. 5, when the processor 610 recognizes the legality value of the first publisher using the advertisement performance value of the first publisher, it may recognize whether the legality value of the first publisher exceeds a threshold value. (S410).

When it is recognized that the first publisher's legality value exceeds the threshold value (S410, Yes), the processor 610 may restrict the first publisher's access to the advertisement network 400 (S420).

Specifically, each of the plurality of publishers may access the advertisement network 400 based on their identification information. When each of the plurality of publishers accesses the advertisement network 400, the processor 610 compares the identification information of each of the plurality of publishers with the identification information of the publisher with limited access to determine which of the plurality of publishers is a publisher with limited access. I can recognize it. When the access to the ad network 400 of a publisher with limited access among a plurality of publishers (for example, login to the ad network, etc.) is recognized, the processor 610 restricts the access of the publisher with limited access (for example, For example, it is possible to perform a login restriction for an advertisement network).

On the other hand, when the access to the ad network of the first publisher is restricted in step S420, the processor 610 may recognize whether a preset time has elapsed from the time when the access to the ad network of the first publisher is restricted. Can be (S430).

If the processor 610 recognizes that a preset time has elapsed from the time when the access to the ad network of the first publisher is restricted (S430, Yes), the access restriction to the ad network 400 of the first publisher may be canceled. There is (S440).

Specifically, when the processor 610 recognizes that a preset time has elapsed from the time when the access to the ad network of the first publisher is restricted, the access to the ad network 400 of the first publisher is restricted (for example, Restrictions on login to the ad network 400) may be canceled to allow the first publisher to access the ad network 400 (eg, log in to the ad network 400).

More specifically, the processor 610 searches for at least one publisher 500 whose access to the advertisement network 400 is restricted at regular time intervals, and if the access restriction time of the corresponding publisher 500 has passed a preset time. The access restriction of the corresponding publisher 500 may be canceled.

According to some embodiments of the present disclosure, at least one publisher 500 whose access restriction to the advertisement network 400 has been revoked may have a legality value initialized with the termination of the access restriction.

On the other hand, if the processor 610 recognizes that the preset time has not elapsed from the time when the access to the ad network of the first publisher is restricted (S430, No), the access to the ad network 400 of the first publisher is restricted. Can keep.

6 is a flowchart illustrating an example of a method of analyzing at least one user's departure by analyzing an action log according to some embodiments of the present disclosure.

Referring to FIG. 6, the processor 610 may obtain at least one user's departure information by analyzing at least one action log recorded in the first blockchain network 200 (S510).

Specifically, the processor 610 may analyze an action log of at least one user received through the communication unit 630 and recognize a user who does not generate a new action log for a predetermined period of time among the at least one user. Further, the processor 610 may analyze the last received action log of the corresponding user to obtain departure information of the corresponding user.

For example, the processor 610 may analyze the last received action log if a new action log of the first user who downloaded the content by clicking the advertisement and played the game has not been generated for more than 10 days. Further, the processor 610 may obtain departure information, which is information on the reason for the departure of the first user. However, it is not limited to the above-described example, and the last received action log may include action logs in various contents such as web sites, real-time streaming, real-time chat, and games.

Meanwhile, the processor 610 may generate a departure prediction model using the acquired departure information (S520).

Specifically, the departure prediction model inputs at least one action log as training data to each of one or more input nodes included in the input layer of one or more network functions, and one or more hidden in the input layer through a link connected to each input node. By propagating through the layers to the output layer, the output of the deviation prediction model can be derived.

In addition, when training a departure prediction model, an error is derived by comparing the data labeled in the training data with the output of the departure prediction model, and the derived error is at least one hidden layer from the output layer of one or more network functions of the departure prediction model. By propagating to the input layer through the network, one or more network functions can be trained by updating the weight set for each link. Hereinafter, a detailed description of the generation of the departure prediction model will be described later in more detail with reference to FIGS. 7 to 8.

Meanwhile, the processor 610 may calculate the action log using the departure prediction model generated in step S520 (S530).

Specifically, an action log value of the first user may be input to the first input node 901 of the input layer 930. The processor 610 may calculate a value input to the input layer 930 of the network function through weights and propagate to the output node 962 included in the output layer 960 through one or more hidden nodes.

As described above, the processor 610 may calculate the departure prediction model using the action log, and a detailed description of the calculation using the departure prediction model will be described later in more detail with reference to FIGS. 7 to 8.

The processor 610 may predict a departure rate and a departure time of each user through the departure data calculated using the departure prediction model (S540).

Specifically, the processor 610 uses the departure data calculated using the churn prediction model, and the churn rate at the current point of time at intervals such as after a certain period of time, after content update, and at a certain interval of all users introduced through the first publisher. And the departure point can be predicted.

7 is a schematic diagram illustrating a network function according to some embodiments of the present disclosure.

Throughout this specification, a model, a neural network, a network function, and a neural network may be used with the same meaning. A neural network may consist of a set of interconnected computational units, which may generally be referred to as “nodes”.

These “nodes” may also be referred to as “neurons”. The neural network includes at least one or more nodes. The nodes (or neurons) that make up neural networks may be interconnected by one or more "links".

In a neural network, one or more nodes connected through a link may relatively form a relationship between an input node and an output node. The concepts of an input node and an output node are relative, and any node in an output node relationship with respect to one node may have an input node relationship in a relationship with another node, and vice versa. As described above, the relationship between the input node and the output node may be created around a link. One or more output nodes may be connected to one input node through a link, and vice versa. In the relationship between the input node and the output node connected through one link, the value of the output node may be determined based on data input to the input node. Here, a node interconnecting the input node and the output node may have a weight.

The weight may be variable and may be changed by a user or an algorithm in order for the neural network to perform a desired function.

For example, when one or more input nodes are interconnected to one output node by respective links, the output node includes values input to input nodes connected to the output node and a link corresponding to each input node. The output node value may be determined based on the weight. As described above, in a neural network, one or more nodes are interconnected through one or more links to form an input node and an output node relationship in the neural network.

The characteristics of the neural network may be determined according to the number of nodes and links in the neural network, the association relationship between the nodes and the links, and a weight value assigned to each of the links.

For example, when the same number of nodes and links exist, and two neural networks having different weight values between the links exist, the two neural networks may be recognized as being different from each other. A neural network can be configured including one or more nodes. Some of the nodes constituting the neural network may configure one layer based on distances from the initial input node.

For example, a set of nodes having a distance n from an initial input node may constitute n layers. The distance from the first input node may be defined by the minimum number of links that must be traversed to reach the node from the first input node.

However, the definition of such a layer is arbitrary for explanation, and the order of the layer in the neural network may be defined in a different way than that described above.

For example, the layer of nodes may be defined by the distance from the last output node. The initial input node may mean one or more nodes to which data is directly input without going through a link in a relationship with other nodes among nodes in the neural network.

Alternatively, in a relationship between nodes in a neural network network based on a link, it may mean nodes that do not have other input nodes connected by a link.

Similarly, the final output node may mean one or more nodes that do not have an output node in a relationship with other nodes among nodes in the neural network.

In addition, the hidden node may refer to nodes constituting a neural network other than the first input node and the last output node.

In the neural network according to some embodiments of the present disclosure, the number of nodes in the input layer may be the same as the number of nodes in the output layer, and the number of nodes decreases and then increases again as the input layer proceeds to the hidden layer. I can.

In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes of an input layer is less than the number of nodes of an output layer, and the number of nodes decreases as the number of nodes progresses from the input layer to the hidden layer. have.

In addition, the neural network according to another exemplary embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer is greater than the number of nodes in the output layer, and the number of nodes increases as the input layer proceeds to the hidden layer. I can.

A neural network according to still another embodiment of the present disclosure may be a neural network in the form of a combination of the aforementioned neural networks.

A deep neural network (DNN) may mean a neural network including a plurality of hidden layers in addition to an input layer and an output layer.

Deep neural networks can be used to identify potential structures in data. In other words, it is possible to grasp the potential structure of photos, texts, videos, voices, and music (e.g., what objects are in photos, what are the content and emotions of the text, what are the contents and emotions of the voice, etc.) .

Deep neural networks include convolutional neural networks (CNNs), recurrentneural networks (RNNs), auto encoders, Generative Adversarial Networks (GANs), and restricted boltzmann machines (RBMs). machine), deep belief network (DBN), Q network, U network, Siam network, and the like.

The foregoing description of the deep neural network is only an example, and the present disclosure is not limited thereto. The neural network may be learned by at least one of supervised learning, unsupervised learning, and semi supervised learning.

Learning of neural networks is to minimize output errors. In learning of a neural network, iteratively inputs training data to the neural network, calculates the output of the neural network for the training data and the error of the target, and reduces the error from the output layer of the neural network to the input layer. This is a process of updating the weight of each node of the neural network by backpropagation in the direction.

In the case of teacher learning, learning data in which the correct answer is labeled in each learning data is used (ie, labeled learning data), and in the case of non-satellite learning, the correct answer may not be labeled in each learning data.

That is, for example, in the case of teacher learning related to data classification, the learning data may be data in which a category is labeled with each learning data. Labeled training data is input to a neural network, and an error may be calculated by comparing an output (category) of the neural network with a label of the training data.

As another example, in the case of comparative history learning regarding data classification, an error may be calculated by comparing the input training data with the neural network output. The calculated error is backpropagated in the neural network in the reverse direction (ie, from the output layer to the input layer), and the connection weight of each node of each layer of the neural network may be updated according to the backpropagation. The amount of change in the connection weight of each node to be updated may be determined according to a learning rate. The computation of the neural network for the input data and the backpropagation of the error can constitute a learning cycle (epoch).

The learning rate may be applied differently according to the number of repetitions of the learning cycle of the neural network. For example, a high learning rate is used in the early stages of learning of a neural network, so that the neural network quickly secures a certain level of performance to increase efficiency, and a low learning rate can be used in the later stages of learning to increase accuracy. In the learning of a neural network, in general, the training data may be a subset of actual data (that is, data to be processed using the learned neural network), and thus, errors in the training data decrease, but errors in the actual data occur. There may be increasing learning cycles.

Overfitting is a phenomenon in which errors in actual data increase due to over-learning on learning data. For example, a neural network learning a cat by showing a yellow cat may not recognize that it is a cat when it sees a cat other than yellow, which may be a kind of overfitting. Overfitting can cause an increase in errors in machine learning algorithms. Various optimization methods can be used to prevent such overfitting. In order to prevent overfitting, methods such as increasing training data, regulation, or dropout in which some nodes of the network are omitted during the training process may be applied.

8 is a diagram for describing a legality numerical calculation model and a departure prediction model according to some embodiments of the present disclosure.

However, at least one or more hidden layers may be included between the hidden layer 1 940 and the hidden layer 2 950, and at least one or more hidden layers may be omitted and illustrated in this exemplary view.

First, a learning method in a legality numerical calculation model according to some embodiments of the present disclosure will be described.

The processor 610 may generate a legality numerical calculation model including at least one network function. The network function may include at least one input layer 930 and at least one output layer 960.

In addition, the processor 610 may input items for learning of calculating the legality numerical value to the network function of the model for calculating the legality numerical value. That is, at least one of a learning advertisement fraud value, a learning ROAS value, and a learning conversion rate may be input to each of the input nodes included in the input layer 930 of the network function of the legality calculation model.

The processor 610 may calculate each item value input to the input node included in the input layer 930 of the network function with a weight set in a link connected to the input node and propagate it to the hidden layer. For example, The first hidden node 920 included in the hidden layer 1 940 is a value transmitted to the first input node 901 and a value obtained by calculating a first weight 911, and a value transmitted to the second input node 902. The value and the value obtained by calculating the second weight, the value transmitted to the third input node 903 and the value obtained by calculating the third weight, the value transmitted to the fourth input node 904 and the value obtained by calculating the fourth weight, and A value transmitted to the fifth input node 905 and a value obtained by calculating the fifth weight may be received.

For example, the first hidden node 921 included in the hidden layer 1 940 is a value obtained by multiplying the value transmitted to the first input node 901 and the first weight 911, and the second input node 902 The value transmitted to the value multiplied by the second weight, the value transmitted to the third input node 903 and the product of the third weight, the value transmitted to the fourth input node 904 and the value obtained by multiplying the fourth weight, and A sum of a value obtained by multiplying a value transmitted to the fifth input node 905 and a fifth weight may be received. However, the method of calculating the above-described weight is only an example and is not limited thereto.

At least one of the learning advertisement fraud value, the learning ROAS value, and the learning conversion rate may be propagated from the input layer 930 of the network function to the output layer 960 through the hidden layer 1 940 and hidden layer 2 950. The weight of the network function may be adjusted based on an error between the legitimacy value (ie, output) that is an output value from the output node 902 included in the output layer 960 and the learning legitimacy value (ie, correct answer). The processor 610 may update weights set for each link while propagating an error from the output layer 960 of the network function to the input layer 930 through one or more hidden layers. Here, updating of the weights may mean that the weight of the W1 (1,1) 911 is adjusted after the weight of the W2(1,1) 931 is adjusted.

As described above, a model for calculating a legality value may be generated according to some embodiments of the present disclosure.

Hereinafter, a learning method in a departure prediction model according to some embodiments of the present disclosure will be described.

The processor 610 may generate a departure prediction model including at least one or more network functions. The network function may include at least one input layer 930 and at least one output layer 960.

Also, the processor 610 may input a learning action log to a network function of the departure prediction model. Each item of the learning action log may be input to each input node included in the input layer 930 of the network function of the deviation prediction model.

For example, the first user's content playback interval is the first input node 901, the first user's purchase interval of the in-game goods is the second input node 902, and the first user's product purchase refund is the third input. The node 903 may be input to the fourth input node 904 to delete the content of the first user, and the fifth input node 905 to reinstall the content of the first user. However, the item value for each node is not limited to the above-described example.

The processor 610 may calculate each item value input to the input node included in the input layer 930 of the network function with a weight set in a link connected to the input node and propagate it to the hidden layer. For example, The first hidden node 920 included in the hidden layer 1 940 is a value transmitted to the first input node 901 and a value obtained by calculating a first weight 911, and a value transmitted to the second input node 902. The value and the value obtained by calculating the second weight, the value transmitted to the third input node 903 and the value obtained by calculating the third weight, the value transmitted to the fourth input node 904 and the value obtained by calculating the fourth weight, and A value transmitted to the fifth input node 905 and a value obtained by calculating the fifth weight may be received. For example, the first hidden node 920 included in the hidden layer 1 940 is a value obtained by multiplying the value transmitted to the first input node 901 and the first weight 911, and the second input node 902 The value transmitted to the value multiplied by the second weight, the value transmitted to the third input node 903 and the product of the third weight, the value transmitted to the fourth input node 904 and the value obtained by multiplying the fourth weight, and A sum of a value obtained by multiplying a value transmitted to the fifth input node 905 and a fifth weight may be received. However, the method of calculating the above-described weight is only an example and is not limited thereto.

The learning action log of the network function may be propagated from the input layer 930 to the output layer 960 through the hidden layer 1 940 and the hidden layer 2 950. The weight of the network function is calculated based on the error of the output value from the output node 902 included in the output layer 960, the departure rate and the departure point (i.e., output) and the learning departure rate and the departure point (i.e., correct answer). Can be adjusted. The processor 610 may update weights set for each link while propagating an error from the output layer 960 of the network function to the input layer 930 through one or more hidden layers. Here, updating of the weights may mean that the weight of the W1 (1,1) 911 is adjusted after the weight of the W2(1,1) 931 is adjusted.

As described above, a departure prediction model may be generated according to some embodiments of the present disclosure.

In addition, a method of generating a legality value using a model for calculating the legality value generated according to some embodiments of the present disclosure will be described below.

The processor 610 may input each item value of items for calculating the legality value to each input node included in the input layer 930 of the network function.

For example, item values (at least one of an advertisement fraud value, an ROAS value, and a conversion rate) of items for calculating the legality value of the first publisher may be input to the first input node 901 of the input layer 930. The processor 610 weights the values input to the input layer 930 of the network function (a first weight (W1(1,1) 911) in this example) and a second weight (W2(1,1) 931 ))) and propagated to the output node 962 included in the output layer 960 through one or more hidden nodes (including hidden layer 1 940 and hidden layer 2 950 in this example) can do. The processor 610 may transmit the legality value, which is an output value of the output node 962, to the first blockchain network 200 through the communication unit 630 or store it in the storage unit 620.

As described above, according to some embodiments of the present disclosure, at least one publisher's legality value may be generated.

Hereinafter, a method of generating a departure rate and a departure prediction time using a departure prediction model according to some embodiments of the present disclosure will be described.

The processor 610 may input each item value of items included in the user's action log to each input node included in the input layer 930 of the network function.

For example, an action log value for the content reproduction interval of the first user may be input to the first input node 901 of the input layer 930. The processor 610 weights the values input to the input layer 930 of the network function (a first weight (W1(1,1) 911) in this example) and a second weight (W2(1,1) 931 ))) and propagated to the output node 962 included in the output layer 960 through one or more hidden nodes (including hidden layer 1 940 and hidden layer 2 950 in this example) can do. The processor 610 may transmit the legality value, which is an output value of the output node 962, through the communication unit 630 or store it in the storage unit 620.

As described above, according to some embodiments of the present disclosure, at least one user's churn rate and departure time point may be generated.

The order of the steps of FIGS. 2 to 8 described above may be changed as necessary, and at least one or more steps and/or configurations may be omitted or added. In addition, the steps and/or configurations described above are only examples of the present disclosure, and the scope of the present disclosure is not limited thereto.

9 exemplarily shows a connection structure of blocks stored in a first blockchain network according to some embodiments of the present disclosure.

Referring to FIG. 9, blocks, which are storage units in the first blockchain network 200, may be connected to each other in a chain form to form a block chain. Referring to FIG. 9, a block may mean an action block described above in FIGS. 1 to 8.

The action blocks 1010 and 1020 may be composed of block headers 1011 and 1021 and transactions 1012 and 1022. The block headers 1011 and 1021 may include, for example, a hash value of a previous block header, a nonce value, and a hash value of a transaction group. The nonce value is a value that a node in the first blockchain network 200 changes to generate a block, and may be used as an input value of a specific hash function along with other values of the block headers 1011 and 1021. When a specific nonce value is used, if the hash value of the block headers 1011 and 1021 is smaller than the predetermined difficulty value (which can be stored in the block headers 1011 and 1021), the corresponding block headers 1011 and 1021 A hash value for can be determined. The hash value of the transaction group may mean a root hash value of data included in the transactions 1012 and 1022.

Since the block hash value of action block #101 (1010) is used as an input value to obtain the block hash value of the block header (1021) of action block #102 (1020), action block #102 (1020) and action block # 101 (1010) may be connected to each other. In addition, in obtaining the block hash value, since the hash value of the transaction group, which is the hash value representing the transactions 1012 and 1022, is used as an input value, if any forgery behavior occurs for the transactions 1012 and 1022, The hash value of the transaction group is changed. In this way, adjacent action blocks can be connected to each other, and information of transactions 1012 and 1022 in the action block can be stored with integrity.

The transactions 1012 and 1022 of the action blocks 1010 and 1020 are transactions 1012 and 1022 (s) included in the action blocks 1010 and 1020 among the issued transactions 1012 and 1022, and the corresponding transaction 1012 When, 1022 is included in the blocks 1010 and 1020, an operation related to the corresponding transaction 1012 and 1022 may be performed in the first blockchain network 200. Here, the transactions 1012 and 1022 may include action log #1, action log #2, action log #3, and action log #4. However, the present invention is not limited thereto, and various pieces of information may be recorded as transactions 1012 and 1022 in blocks 1010 and 1020.

As an additional embodiment, although not shown in FIG. 9, the block headers 1011 and 1021 of the action blocks 1010 and 1020 are the hash values (ParentHash) of the parent block (previous block), the uncle block of the current block (block difficulty level). Hash values (UncleHash) of blocks that were not adopted as blocks because of the relatively low value (UncleHash), 32-byte Bloom filter information (Bloom) used to use log information, the difficulty of the previous block and the block difficulty (Difficulty) calculated by timestamp, A value used to ensure the number of calculations sufficient to calculate the hash value from the current block number (Number), the initial creation time (Time) of the block, other information (Extra) of the block, and/or proof of work (Mix Digest, Nonce) It may also include. However, the present invention is not limited thereto, and the block headers 1011 and 1021 may further include data or information other than those described above.

The transaction of the action blocks 1010 and 1020 is the transaction(s) included in the block among issued transactions, and when the transaction is included in the block, an operation related to the transaction is performed in the first blockchain network 200 Can be.

In addition, the block header 1011 of the action block #101 (1010) may include a first Merkle root value 1130 for information stored in each transaction of at least one action block. The root value 1130 will be described.

10 is an exemplary diagram for describing an example of a method of generating a first Merkle root value according to some embodiments of the present invention.

Referring to FIG. 10, when an action node generates a block, a first Merkle root value 1130 may be generated by using a plurality of action logs to be recorded as transactions of a block to be generated.

Specifically, the action node receives the action log #1 (1111) from among the plurality of action logs to be recorded as a transaction of the block to be created, through a hash algorithm (eg, Secure Hash Algorithm-256, etc.) to the first hash value (1121). Can be converted to In addition, the action node may convert action log #2 1112 among a plurality of action logs to be recorded as a transaction of a block to be generated into a second hash value 1122 through a hash algorithm. In addition, the action node may combine the first hash value 1121 and the second hash value 1122 and generate the first Merkle root value 1130 through a hash algorithm.

The first Merkle root value 1130 generated as described above may be recorded in a block header of a block including action log #1 (1111) and action log #2 (1112) as transactions. Therefore, if any one of the plurality of action logs included in the transaction is altered, a Merkle root value completely different from the first Merkle root value 1130 is generated. I can.

According to some embodiments of the present disclosure, the action block may be generated at a specific time interval (eg, 5 minutes, 1 minute), and the specific time may be changed based on the difficulty of generating the action block.

11 exemplarily shows a connection structure of blocks stored in a second blockchain network according to some embodiments of the present disclosure.

Referring to FIG. 11, blocks, which are storage units in the second block chain network 300, may be connected to each other in a chain form to form a block chain. Referring to FIG. 11, the block may mean the legality numeric recording block described above in FIGS. 1 to 8.

The legality numeric recording blocks 1010 and 1020 may be composed of block headers 1211 and 1221 and transactions 1212 and 1222. The block headers 1211 and 1221 may include, for example, a hash value of a previous block header, a nonce value, and a hash value of a transaction group. The nonce value is a value that a node in the second blockchain network 300 changes to generate a block, and may be used as an input value of a specific hash function along with other values of the block headers 1211 and 1221. When a specific nonce value is used, if the hash value of the block headers 1211 and 1221 is smaller than the predetermined difficulty value (which can be stored in the block headers 1211 and 1221), the corresponding block headers 1211 and 1221 A hash value for can be determined. The hash value of the transaction group may mean a root hash value of data included in the transactions 1212 and 1222.

Since the block hash value of the legality numeric recording block #201 (1210) is used as an input value to obtain the block hash value of the block header 1221 of the legality numeric recording block 202 (1220), the legality numeric recording block #202 (1220) ) And the legality number recording block #201 (1210) may be connected to each other. In addition, in obtaining the block hash value, since the hash value of the transaction group, which is the hash value representing the transactions 1212 and 1222, is used as an input value, if any forgery behavior occurs for the transactions 1212 and 1222, The hash value of the transaction group is changed. In this way, adjacent blocks may be connected to each other, and information of transactions 1212 and 1222 in the block may be stored with integrity.

The transactions 1212 and 1222 of the legality numeric recording blocks 1210 and 1220 are among the issued transactions 1212 and 1222 as the transaction 1212 and 1222 (s) included in the legality numeric recording blocks 1210 and 1220, When the corresponding transactions 1212 and 1222 are included in the legality numeric recording blocks 1210 and 1220, an operation related to the corresponding transactions 1212 and 1222 may be performed in the second blockchain network 300. Here, the transactions 1212 and 1222 may include the legality value of the first publisher, the legality value of the second publisher, the legality value of the third publisher, and the legality value of the fourth publisher. However, the present disclosure is not limited thereto, and various pieces of information may be recorded as transactions 1212 and 1222 in the legality numeric recording blocks 1210 and 1220.

As an additional embodiment, although not shown in FIG. 11, the block headers 1211 and 1221 of the legality numeric recording blocks 1210 and 1220 are a hash value (ParentHash) of a parent block (previous block), an uncle block of the current block (block The hash values (UncleHash) of blocks that were not adopted as blocks due to the relatively low difficulty of the block), 32-byte bloom filter information used to use log information (Bloom), the difficulty of the previous block and the block difficulty calculated by the timestamp (Difficulty) ), the current block number (Number), the initial creation time of the block (Time), other information of the block (Extra), and/or a value used to ensure the number of calculations sufficient to calculate the hash value from the proof of work (Mix Digest, Nonce) may be included. However, the present invention is not limited thereto, and the block headers 1211 and 1221 may further include data or information other than those described above.

The transaction of the legality numerical record blocks 1210 and 1220 is the transaction(s) included in the block among issued transactions, and when the transaction is included in the block, the operation related to the transaction is performed by the second blockchain network 300 Can be done in

In addition, the block header 1211 of the legality number recording block #201 (1210) is a second Merkle root value 1330 for the legality value of the first publisher and the legality value of at least one other publisher 500 other than the first publisher. ), and the second Merkle root value 1330 will be described with reference to FIG. 12 below.

12 is an exemplary diagram illustrating an example of a method of generating a second Merkle root value according to some embodiments of the present invention.

Referring to FIG. 12, when generating a block, the legality number recording node may generate a second Merkle root value 1330 by using a plurality of legality values to be recorded as a transaction of a block to be generated.

Specifically, the legality number recording node uses a hash algorithm (for example, Secure Hash Algorithm-256, etc.) to a third hash of the legality value (1311) of the first publisher among the plurality of legality values to be recorded as a transaction of the block to be created. It can be converted to a value 1321. In addition, the legality number recording node may convert the second publisher's legality value 1312 from among a plurality of legality values to be recorded as a transaction of the block to be generated into a fourth hash value 1322 through a hash algorithm. In addition, the legality numeric recording node may combine the third hash value 1321 and the fourth hash value 1322 and generate a second Merkle root value 1330 through a hash algorithm.

The second Merkle root value 1330 generated as described above may be recorded in a block header of a block including the legality value 1311 of the first publisher and the legality value 1312 of the second publisher as a transaction. Therefore, if any one of the plurality of legality values included in the transaction is altered, a Merkle root value completely different from the second Merkle root value 1330 is generated. Therefore, it is necessary to verify the integrity of the plurality of action logs included in the transaction. I can.

According to some embodiments of the present disclosure, the legality numeric recording block may be generated at a specific time interval (eg, 5 minutes, 1 minute), wherein the specific time may be changed based on the difficulty of generating the legality numeric recording block. I can.

13 shows a general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure may be implemented.

While the present disclosure has generally been described above with respect to computer-executable instructions that may be executed on one or more computers, those skilled in the art will appreciate that the present disclosure may be implemented in combination with other program modules and/or as a combination of hardware and software. I will know.

In general, modules herein include routines, procedures, programs, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Further, to those skilled in the art, the method of the present disclosure is not limited to single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable household appliances, and the like (each of which It will be appreciated that other computer system configurations may be implemented, including one or more associated devices).

The described embodiments of the present disclosure may also be practiced in a distributed computing environment where certain tasks are performed by remote processing devices that are connected through a communication network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer-readable media. A computer-accessible medium includes volatile and non-volatile media, transitory and non-transitory media, and removable and non-removable media. By way of example, and not limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media.

Computer-readable storage media include volatile and nonvolatile media, transitory and non-transitory media, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Includes the medium. Computer-readable storage media include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage, or other magnetic storage. Devices, or any other medium that can be accessed by a computer and used to store desired information.

Computer-readable transmission media typically implement computer-readable instructions, data structures, program modules or other data on a modulated data signal such as a carrier wave or other transport mechanism. Includes all information delivery media. The term modulated data signal refers to a signal in which one or more of the characteristics of the signal is set or changed to encode information in the signal. By way of example, and not limitation, computer-readable transmission media include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above-described media are also intended to be included within the scope of computer-readable transmission media.

An exemplary environment 1400 is shown that implements various aspects of the present disclosure, including a computer 1402, which includes a processing unit 1404, a system memory 1406, and a system bus 1408. do. System bus 1408 couples system components, including but not limited to, system memory 1406 to processing device 1404. The processing unit 1404 may be any of a variety of commercially available processors. Dual processors and other multiprocessor architectures may also be used as processing unit 1404.

The system bus 1408 may be any of several types of bus structures that can be additionally interconnected to a memory bus, a peripheral bus, and a local bus using any of a variety of commercial bus architectures. System memory 1406 includes read-only memory (ROM) 1410 and random access memory (RAM) 1412. The basic input/output system (BIOS) is stored in nonvolatile memory 1410 such as ROM, EPROM, EEPROM, etc. This BIOS is a basic input/output system that helps transfer information between components in the computer 1402, such as during startup. Includes routines. RAM 1412 may also include high speed RAM such as static RAM for caching data.

Computer 1402 also includes internal hard disk drive (HDD) 1414 (e.g., EIDE, SATA)—this internal hard disk drive 1414 can also be configured for external use within a suitable chassis (not shown). Yes--, magnetic floppy disk drive (FDD) 1416 (e.g., to read from or write to removable diskette 1418), and optical disk drive 1420 (e.g., CD-ROM For reading the disk 1422 or reading from or writing to other high-capacity optical media such as DVD). The hard disk drive 1414, the magnetic disk drive 1416 and the optical disk drive 1420 are each connected to a system bus 1408 by a hard disk drive interface 1424, a magnetic disk drive interface 1426 and an optical drive interface 1428, respectively. ) Can be connected. The interface 1424 for implementing an external drive includes, for example, at least one or both of USB (Universal Serial Bus) and IEEE 1394 interface technologies.

These drives and their associated computer readable media provide non-volatile storage of data, data structures, computer executable instructions, and the like. In the case of computer 1402, drives and media correspond to storing any data in a suitable digital format. Although the description of the computer-readable storage medium above refers to a removable optical medium such as a HDD, a removable magnetic disk, and a CD or DVD, those skilled in the art may include a zip drive, a magnetic cassette, a flash memory card, a cartridge, It will be appreciated that other types of computer-readable storage media, such as etc., may also be used in the exemplary operating environment and that any such media may contain computer-executable instructions for performing the methods of the present disclosure. .

A number of program modules may be stored in the drive and RAM 1412, including the operating system 1430, one or more application programs 1432, other program modules 1434, and program data 1436. All or part of the operating system, applications, modules, and/or data may also be cached in RAM 1412. It will be appreciated that the present disclosure may be implemented on a number of commercially available operating systems or combinations of operating systems.

A user may input commands and information to the computer 1402 through one or more wired/wireless input devices, for example, a pointing device such as a keyboard 1438 and a mouse 1440. Other input devices (not shown) may include a microphone, IR remote control, joystick, game pad, stylus pen, touch screen, and the like. These and other input devices are often connected to the processing unit 1404 through the input device interface 1442, which is connected to the system bus 1408, but the parallel port, IEEE 1394 serial port, game port, USB port, IR interface, It can be connected by other interfaces such as etc.

A monitor 1444 or other type of display device is also connected to the system bus 1408 through an interface, such as a video adapter 1446. In addition to the monitor 1444, the computer generally includes other peripheral output devices (not shown) such as speakers, printers, etc.

Computer 1402 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1448, via wired and/or wireless communication. The remote computer(s) 1448 may be a workstation, server computer, router, personal computer, portable computer, microprocessor-based entertainment device, peer device, or other common network node, and is generally referred to as computer 1402. It includes many or all of the described components, but for simplicity, only memory storage device 1450 is shown. The logical connections shown include wired/wireless connections to a local area network (LAN) 1452 and/or to a larger network, eg, a wide area network (WAN) 1454. Such LAN and WAN networking environments are common in offices and companies, and facilitate an enterprise-wide computer network such as an intranet, all of which can be connected to a worldwide computer network, for example the Internet.

When used in a LAN networking environment, the computer 1402 is connected to the local network 1452 via a wired and/or wireless communication network interface or adapter 1456. Adapter 1456 may facilitate wired or wireless communication to LAN 1452, which LAN 1452 also includes a wireless access point installed therein to communicate with wireless adapter 1456. When used in a WAN networking environment, the computer 1402 may include a modem 1458, connected to a communication server on the WAN 1454, or through the Internet, etc. to establish communication through the WAN 1454. Have means. Modem 1458, which may be an internal or external and wired or wireless device, is connected to the system bus 1408 through a serial port interface 1442. In a networked environment, program modules described for computer 1402 or portions thereof may be stored in remote memory/storage device 1450. It will be appreciated that the network connections shown are exemplary and other means of establishing communication links between computers may be used.

Computer 1402 is associated with any wireless device or entity deployed and operated in wireless communication, e.g., a printer, scanner, desktop and/or portable computer, portable data assistant (PDA), communication satellite, wireless detectable tag. It operates to communicate with any equipment or place and phone. This includes at least Wi-Fi and Bluetooth wireless technologies. Thus, the communication may be a predefined structure as in a conventional network or may simply be ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) allows you to connect to the Internet, etc. without wires. Wi-Fi is a wireless technology such as a cell phone that allows such devices, for example computers, to transmit and receive data indoors and outdoors, ie anywhere within the coverage area of a base station. Wi-Fi networks use a wireless technology called IEEE 802.14 (a, b, g, etc.) to provide secure, reliable and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to a wired network (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in unlicensed 2.4 and 5 GHz wireless bands, for example at a data rate of 14 Mbps (802.14a) or 54 Mbps (802.14b), or in products that include both bands (dual band). have.

A person of ordinary skill in the art of the present disclosure includes various exemplary logical blocks, modules, processors, means, circuits and algorithm steps described in connection with the embodiments disclosed herein, electronic hardware, (convenience). For the sake of clarity, it will be appreciated that it may be implemented by various forms of program or design code or a combination of both (referred to herein as "software"). To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the particular application and design constraints imposed on the overall system. A person of ordinary skill in the art of the present disclosure may implement the described functions in various ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

The various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and/or engineering techniques. The term “article of manufacture” includes a computer program, carrier, or media accessible from any computer-readable device. For example, computer-readable storage media include magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CD, DVD, etc.), smart cards, and flash Memory devices (eg, EEPROM, cards, sticks, key drives, etc.), but are not limited thereto. The term “machine-readable medium” includes, but is not limited to, wireless channels and various other media capable of storing, holding, and/or transmitting instruction(s) and/or data.

It is to be understood that the specific order or hierarchy of steps in the presented processes is an example of exemplary approaches. Based on the design priorities, it is to be understood that within the scope of the present disclosure a specific order or hierarchy of steps in processes may be rearranged. The appended method claims provide elements of the various steps in a sample order, but are not meant to be limited to the specific order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those of ordinary skill in the art, and general principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present disclosure is not limited to the embodiments presented herein, but is to be interpreted in the widest scope consistent with the principles and novel features presented herein.

Claims (23)

A computer program stored on a computer-readable storage medium, comprising:
The computer program includes instructions for causing the server to perform the following operations, the operations:
Recognizing an action log of at least one user from at least one action block recorded in the first blockchain network;
Calculating an advertisement performance value of a first publisher among a plurality of publishers using the action log;
By transmitting the advertisement performance value of the first publisher to at least one action node included in the first blockchain network, a plurality of action nodes verify the validity of the advertisement performance value through a consensus algorithm, and the advertisement performance Causing the plurality of action nodes to record the advertisement performance values in a transaction in an action block as the validity of the values is verified;
Recognizing a legality value of the first publisher based on the advertisement performance value of the first publisher;
When the legality value of the first publisher is recognized, the legality value is transmitted to at least one legality value recording node included in the second blockchain network, so that a plurality of legality number recording nodes can use the consensus algorithm. Verifying the validity of, and causing the plurality of legality number recording nodes to record the legality value as a transaction in the legality number recording block as the validity of the legality value is verified; And
Restricting access to the advertisement network of the first publisher when the legality value of the first publisher exceeds a preset threshold value;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
Recognizing the action log of at least one user in the at least one action block recorded in the first blockchain network,
Dividing each of the at least one user into a genuine user and a non-genuine user by using the action log after recognizing the action log;
Further comprising,
A computer program stored on a computer-readable storage medium. The method of claim 2,
The operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log,
Calculate a return on advertising spend (ROAS) value by using the number of genuine users, profits according to the genuine users, the number of non-genuine users, profits according to the non-genuine users, and a first weight value set for the genuine users Action;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 2,
The operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log,
The number of genuine users, profits according to the genuine users, the number of non-genuine users, the profits according to the non-genuine users, a first weight value set for the genuine user, and a second weight value set for the non-genuine users are used. Calculating the ROAS value;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 2,
The operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log,
Calculating a conversion rate using the number of clicks of the advertisement by the genuine user, the number of clicks of the advertisement by the non-genuine user, and a third weight value set for the genuine user;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 2,
The operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log,
Calculating a conversion rate using the number of clicks of the advertisement by the genuine user, the number of clicks of the advertisement by the non-genuine user, a third weight value set for the genuine user, and a fourth weight value set for the non-genuine user;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
The operation of calculating the advertisement performance value of the first publisher among the plurality of publishers using the action log,
Calculating an advertisement fraud value based on first information on the number of suspicious reports of the first publisher included in the action log;
Including,
The first information,
Including at least one of the number of suspicious reports on the number of advertisements by the first publisher, the number of suspicious reports on the advertisement content, and the number of suspicious reports on the advertisement period,
A computer program stored on a computer-readable storage medium. The method of claim 2,
The action log of the at least one user,
Including at least one of user identification information for each of the at least one user, information on the time at which the event occurred, information on the terminal where the event was executed, information on payment of content related to the event, and usage of the content related to the event and,
The operation of dividing each of the at least one user into a genuine user and a non-genuine user using the action log,
The at least one of the user identification information, the time information at which the event occurred, information on the terminal on which the event was executed, the content payment information related to the event, and the usage amount of the content related to the event Dividing each user into a genuine user and a non-genuine user;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 8,
The above event is:
Including at least one of an advertisement click event, an application installation event, and a preset action event,
A computer program stored on a computer-readable storage medium. The method of claim 1,
The operation of transmitting the action log to at least one action node included in the first blockchain network,
Generating a first transaction signed with a private key of the advertisement management server; And
Transmitting the generated first transaction to the at least one action node;
It further includes,
The first transaction,
A first transaction input value including identification information generated from the private key; And
A first transaction output value including address information in the first blockchain network in which the action log and the first transaction including the action log are to be stored;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
Recognizing the legality value of the first publisher based on the advertisement performance value of the first publisher,
Calculating the legality value by calculating at least one of an advertisement fraud value, an ROAS value, and a conversion rate included in the advertisement performance value using a legality value calculation model;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 11,
The legality numerical calculation model,
A training data set including at least one of a learning advertisement fraud value, a learning ROAS value, and a learning conversion rate is input to each of one or more input nodes included in the input layer of one or more network functions, and through a link connected to each of the input nodes. By propagating from the input layer to the output layer through one or more hidden layers, the output of the legality numerical calculation model is derived, and an error is calculated by comparing the learning legality numerical value labeled in each of the training data sets with the output of the legality numerical calculation model. The at least one network by deriving and propagating the derived error from the output layer of at least one network function of the legitimacy numerical calculation model to the input layer through the at least one hidden layer to update the weights set for each link. A model created by learning a function,
A computer program stored on a computer-readable storage medium. The method of claim 1,
Collecting a new action log for each of at least one user at a preset time interval; And
By transmitting the new action log to the at least one action node included in the first blockchain network, when the new action log is verified through a consensus algorithm, a new action block is generated in each of the plurality of action nodes. Causing the new action log to be recorded in the new action block;
Further comprising,
A computer program stored on a computer-readable storage medium. The method of claim 13,
The operation of transmitting the new action log to at least one action node included in the first blockchain network,
Generating a second transaction signed with a private key of the advertisement management server; And
Transmitting the generated second transaction to the at least one action node;
Including,
The second transaction,
A second transaction input value including identification information generated from the private key; And
A second transaction output value including address information in the first blockchain network in which the new action log and the second transaction including the new action log are to be stored;
Containing,
A computer program stored on a computer-readable storage medium. delete The method of claim 1,
When the legality value of the first publisher is recognized, the operation of transmitting the legality value to at least one legality value recording node included in the second blockchain network,
Generating a third transaction signed with a private key of the advertisement management server; And
Transmitting the generated third transaction to the at least one action node;
Including,
The third transaction,
A third transaction input value including identification information generated from the private key; And
A third transaction output value including address information in the second blockchain network in which the legality value and the third transaction including the legality value are to be stored;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
Recognizing the legality value of the first publisher based on the advertisement performance value of the first publisher,
Recognizing a legality value of the first publisher when there is a request for access to the advertising network by the first publisher;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
When the legality value of the first publisher exceeds a preset threshold, the operation of restricting access to the advertising network of the first publisher,
Restricting access of the first publisher to the advertisement network for a preset time from when it is recognized that the legality value of the first publisher exceeds the preset threshold value; And
When the preset time has elapsed, canceling the restriction on the access of the first publisher to the advertisement network;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
Analyzing the action log to obtain departure information of the at least one user;
Generating a departure prediction model using the departure information; And
Predicting a user churn rate and a departure time point of each of the at least one user by calculating the action log using the departure prediction model;
Further comprising,
A computer program stored on a computer-readable storage medium. The method of claim 19,
The operation of generating a departure prediction model using the departure information,
Input the action log as training data to each of one or more input nodes included in the input layer of one or more network functions, and propagate from the input layer to the output layer through one or more hidden layers through a link connected to each of the input nodes. Thereby deriving an output of the departure prediction model; And
An error is derived by comparing the data labeled in the training data with the output of the departure prediction model, and the derived error is input from the output layer of one or more network functions of the departure prediction model through the one or more hidden layers. Learning the one or more network functions by updating a weight set for each link by propagating to a layer;
Containing,
A computer program stored on a computer-readable storage medium. The method of claim 1,
The block header of each of the at least one action block,
Including a first Merkle root value for information stored in the transaction of each of the at least one action block,
A computer program stored on a computer-readable storage medium. The method of claim 1,
The block header of the legality numeric recording block,
Comprising a second Merkle root value for the legality value of the first publisher and the legality value of at least one other publisher other than the first publisher,
A computer program stored on a computer-readable storage medium. delete

Images