A method, system, electronic device, and readable storage medium for multi-channel attribution.

Abstract

This invention provides a method, system, electronic device, and readable storage medium for multi-channel attribution. The method includes: when a user registers an account through a device, obtaining and storing a mapping relationship between the registered user ID and the channel ID; attributing the device using multiple attribution algorithms to obtain multiple initial attribution results; determining a target attribution result from the multiple initial attribution results according to a priority rule; and determining whether the target attribution result is correct based on the mapping relationship. This aims to improve the accuracy of channel attribution.

Description

Technical Field

[0001] This invention relates to the field of channel attribution technology, and in particular to a method, system, electronic device, and readable storage medium for multi-channel attribution. Background Technology

[0002] With the rapid development of the internet, all kinds of applications and apps have emerged one after another. People are now used to getting information from Douyin, Toutiao, Kuaishou, and other similar platforms. As users' demand for mobile apps grows, businesses, companies, and individuals from all walks of life are entering the app development field. Once an app is developed, it needs to be promoted so that more people can learn about it, download it, and use it.

[0003] Currently, companies promote their apps through various channels. Determining the source of users from a particular channel involves attributing traffic to that channel. Attribution is generally considered to refer to the effect of mobile app installation and activation being brought about by that channel. Accurate attribution is crucial for optimizing user acquisition and channel selection. In practice, channel package attribution is prone to errors because downloads of channel packages are easily intercepted by mobile app vendors, making it difficult to accurately assess attribution value. Device ID attribution can also lead to errors because users click on informational ads but don't download the app initially, and then download it via an H5 download page. Fuzzy attribution is further complicated by inconsistent IP addresses across different environments, resulting in inaccurate attribution. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, system, electronic device, and readable storage medium for multi-channel attribution, aiming to improve the accuracy of channel attribution.

[0005] The first aspect of this invention provides a method for multi-channel attribution, the method comprising:

[0006] When a user registers an account through a device, the mapping relationship between the registered user ID and the channel ID is obtained and stored;

[0007] The device is attributed using multiple attribution algorithms to obtain multiple initial attribution results;

[0008] Based on priority rules, a target attribution result is determined from the plurality of initial attribution results;

[0009] Based on the mapping relationship, determine whether the target attribution result is correct.

[0010] Optionally, the step of obtaining and storing the mapping relationship between the registered user ID and the channel ID when the user registers an account through the device includes:

[0011] The channel ID of the advertising delivery channel is preset through the event tracking logs;

[0012] When a user registers an account by clicking on an ad through a device, obtain the channel ID and the registered user ID recorded in the ad's event logs;

[0013] Based on the obtained channel ID and registered user ID, create and store the corresponding mapping relationship.

[0014] Optionally, the multiple attribution algorithms include at least the device number attribution algorithm, the channel packet attribution algorithm, and the IP and UA fuzzy attribution algorithm;

[0015] The step of determining the target attribution result from the plurality of initial attribution results according to the priority rule includes:

[0016] The various attribution algorithms are prioritized to obtain corresponding priority rules. In the priority rules, the device number attribution algorithm is the first priority, the IP and UA fuzzy attribution algorithm is the second priority, and the channel package attribution algorithm is the third priority.

[0017] Based on the priority rule, determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results;

[0018] The initial attribution result corresponding to the highest priority attribution algorithm is determined as the target attribution result for the device.

[0019] Optionally, determining the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rule includes:

[0020] The initial attribution result determined by the device number attribution algorithm and the initial attribution result determined by the IP and UA fuzzy attribution algorithm are combined to remove duplicates, and a first deduplication result is obtained.

[0021] The first deduplication result is combined with the initial attribution result corresponding to the channel package attribution algorithm to perform deduplication, and the initial attribution result corresponding to the attribution algorithm with the highest priority among the multiple initial attribution results is determined.

[0022] Optionally, determining whether the target attribution result is correct based on the mapping relationship includes:

[0023] When the channel ID in the target attribution result is the same as the channel ID corresponding to the registered user ID in the mapping relationship, the target attribution result is determined to be correct.

[0024] If the channel ID in the target attribution result is different from the channel ID corresponding to the registered user ID in the mapping relationship, the target attribution result is determined to be incorrect.

[0025] Optionally, the method further includes:

[0026] Attribution precision and attribution recall are calculated based on whether the target attribution results of each device are correct.

[0027] Optionally, attribution accuracy is calculated based on whether the target attribution results of each device are correct, to obtain attribution accuracy, including:

[0028] The number of correct attributions for the target attribution result is determined as the first parameter;

[0029] The number of devices with account logins is determined as the second parameter;

[0030] The attribution accuracy is obtained by dividing the first parameter by the second parameter.

[0031] Optionally, attribution recall is calculated based on whether the target attribution results for each device are correct, to obtain attribution recall, including:

[0032] The number of devices that log in within a time interval, have a mapping relationship, and have a target attribution result is determined as the third parameter;

[0033] The number of devices that download within a preset time range and log in to an account within that time range, and that account has a mapping relationship, is determined as the fourth parameter;

[0034] The attribution recall rate is obtained by dividing the third parameter by the fourth parameter.

[0035] Optionally, the methods for placing ads may include at least mandatory registration and download-based methods.

[0036] A second aspect of the present invention provides a multi-channel attribution system, the system comprising:

[0037] The mapping relationship determination module is used to obtain and store the mapping relationship between the registered user ID and the channel ID when a user registers an account through a device;

[0038] The attribution module is used to attribute the device using multiple attribution algorithms to obtain multiple initial attribution results;

[0039] The target attribution result determination module is used to determine the target attribution result from the plurality of initial attribution results according to priority rules;

[0040] The target attribution result accuracy determination module is used to determine whether the target attribution result is correct based on the mapping relationship.

[0041] Optionally, the mapping relationship determination module includes:

[0042] The Channel ID tracking module is used to preset the channel ID of the advertising delivery channel through tracking logs;

[0043] The Channel ID and Registered User ID Acquisition Module is used to obtain the Channel ID and Registered User ID recorded in the event tracking log of an ad when a user registers an account by clicking on an ad through a device.

[0044] The mapping relationship determination submodule is used to create and store the corresponding mapping relationship based on the obtained channel ID and registered user ID.

[0045] Optionally, the multiple attribution algorithms include at least the device number attribution algorithm, the channel packet attribution algorithm, and the IP and UA fuzzy attribution algorithm;

[0046] The target attribution result determination module includes:

[0047] The priority sorting module is used to sort the various attribution algorithms by priority and obtain the corresponding priority rules. In the priority rules, the device number attribution algorithm is the first priority, the IP and UA fuzzy attribution algorithm is the second priority, and the channel package attribution algorithm is the third priority.

[0048] The first target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rules;

[0049] The second target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm as the target attribution result of the device.

[0050] Optionally, the system further includes:

[0051] The precision and recall determination module is used to calculate attribution precision and attribution recall based on whether the target attribution results of each device are correct, and obtain attribution precision and attribution recall.

[0052] A third aspect of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0053] Memory, used to store computer programs;

[0054] When a processor executes a program stored in memory, it implements the steps of the multi-channel attribution method described in the first aspect of the present invention.

[0055] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the program is executed by a processor, it implements a multi-channel attribution method as described in the first aspect of the present invention.

[0056] The embodiments of the present invention have the following advantages:

[0057] This invention provides a multi-channel attribution method. When a user registers an account through a device, the method acquires and stores the mapping relationship between the registered user ID and the channel ID. Multiple attribution algorithms are used to attribute the device, resulting in multiple initial attribution results. A target attribution result is determined from these initial attribution results based on a priority rule. The mapping relationship related to the registered user ID, stored in a database, is determined based on the registered user ID logged into the target attribution result. If the channel ID in this mapping relationship matches the attribution channel ID in the target attribution result, the target attribution result is deemed correct, thereby improving the accuracy of channel attribution. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a flowchart illustrating a multi-channel attribution method according to an embodiment of the present invention;

[0060] Figure 2 This is a flowchart illustrating multiple attribution algorithms in a multi-channel attribution method according to an embodiment of the present invention;

[0061] Figure 3 This is a schematic diagram illustrating an advertising delivery method in a multi-channel attribution method according to an embodiment of the present invention;

[0062] Figure 4 This is a schematic diagram of a multi-channel attribution system according to an embodiment of the present invention. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] In this invention, Figure 1 This is a flowchart illustrating a multi-channel attribution method according to an embodiment of the present invention. (Refer to...) Figure 1 The present invention provides a multi-channel attribution method, comprising:

[0065] Step S11: When a user registers an account through a device, obtain and store the mapping relationship between the registered user ID and the channel ID;

[0066] Step S12: Attribute the device using multiple attribution algorithms to obtain multiple initial attribution results;

[0067] Step S13: Determine the target attribution result from the plurality of initial attribution results according to the priority rule;

[0068] Step S14: Determine whether the target attribution result is correct based on the mapping relationship.

[0069] In embodiments of the present invention, when promoting the developed application, the company will promote the application through multiple channels. Users can register, download, and install the application through multiple channels. Accurate channel attribution (i.e., determining which channel a user first encountered and began using the application) is crucial for the company to accurately evaluate the promotional effectiveness of each channel, enabling more targeted promotion of the application in the future. Step S11: When the company promotes the developed application through advertisements on various channels, when a user clicks on an advertisement on a channel to register an account via a device (such as a mobile phone or iPad), the channel ID of the channel where the advertisement is located is obtained, and the registered user ID of the user's registered account is recorded. A mapping relationship between the registered user ID and the channel ID is created based on the obtained registered user ID and channel ID, and this mapping relationship is stored in the database. The registered user ID is unique; that is, a registered user ID can only appear once in the registered user directory of an application. If a user subsequently tries to register with a registered user ID already existing in the registered user directory, the system will prompt that the registered user ID already exists and ask the user to register with a different registered user ID.

[0070] For example, a company develops an application and promotes it on channels A, B, and C. User 1 sees the application's promotional ad on channel A and completes the registration of an account with user ID 123ABC by clicking the ad. At the same time, User 2 sees the application's promotional ad on channel B and completes the registration of an account with user ID 1234ABCD by clicking the ad. At this point, a mapping relationship will be created between user ID 123ABC and channel ID A, and a mapping relationship will also be created between user ID 1234ABCD and channel ID B.

[0071] In an embodiment of the present invention, after a user downloads, installs, and logs in to an application via a device, the device is channel-attributed using multiple attribution algorithms. Each attribution algorithm corresponds to an initial attribution result. After obtaining multiple initial attribution results corresponding to the various attribution algorithms, a target attribution result is determined from the multiple initial attribution results corresponding to the various attribution algorithms according to the priority rules of the multiple attribution algorithms. This target attribution result is used as the attribution result for the device. Then, based on the registered user ID of the account logged into by the device, the mapping relationship related to the registered user ID in the mapping relationship stored in the database is determined. The correctness of the target attribution result is determined based on the determined mapping relationship.

[0072] In this invention, determining whether the target attribution result is correct based on the mapping relationship includes: determining that the target attribution result is correct when the channel ID in the target attribution result is the same as the channel ID corresponding to the registered user ID in the mapping relationship; and determining that the target attribution result is incorrect when the channel ID in the target attribution result is different from the channel ID corresponding to the registered user ID in the mapping relationship.

[0073] In an embodiment of the present invention, after determining the mapping relationship related to the registered user ID in the mapping relationship stored in the database, the channel ID in the mapping relationship is compared with the attribution channel ID in the target attribution result. If the two are the same, the target attribution result is determined to be correct; if the two are different, the target attribution result is determined to be incorrect.

[0074] For example, continuing with the above example, when user 1 logs in through a device with the registered user ID 123ABC, the device is channel-attributed using various attribution algorithms, and the channel ID in the corresponding target attribution result is channel B. At this time, the mapping relationship corresponding to the registered user ID 123ABC is obtained from the database. In this mapping relationship, the channel ID corresponding to the registered user ID 123ABC is channel A, while the channel ID in the corresponding target attribution result is channel B. Therefore, it is determined that the target attribution result is incorrect, and the target attribution result cannot be used as a valid attribution result to evaluate the promotion effect of each channel. When User 2 logs in via device with registered user ID 1234ABCD, the device is channel-attributed using various attribution algorithms. The channel ID in the target attribution result is Channel B. At this point, the mapping relationship corresponding to registered user ID 1234ABCD in the database is obtained. In this mapping relationship, the channel ID corresponding to registered user ID 1234ABCD is Channel B. Since the channel ID in the target attribution result is Channel B, it is determined that the target attribution result is correct. This target attribution result can be used as a valid attribution result to evaluate the promotion effect of each channel.

[0075] In this invention, the step of obtaining and storing the mapping relationship between the registered user ID and the channel ID when a user registers an account through a device includes: presetting the channel ID of the advertising delivery channel through the event tracking log; obtaining the channel ID and the registered user ID recorded in the event tracking log of the advertisement when the user clicks on the advertisement to register an account through the device; and creating and storing the corresponding mapping relationship based on the obtained channel ID and the registered user ID.

[0076] In an embodiment of the present invention, one method for obtaining and storing the mapping relationship between the registered user ID and the channel ID when a user registers an account through a device is as follows: When advertising is delivered through various channels, event tracking is performed in each advertisement, and the corresponding channel ID is added to the event tracking log for the channel to which the advertisement is delivered. When a user clicks on an advertisement to register an account through a device, the event tracking log of that advertisement is obtained, and the channel ID of the channel to which the advertisement is delivered is obtained from the event tracking log. At the same time, when the user registers an account, the registered user ID is obtained, and the registered user ID and the channel ID are associated to create a corresponding mapping relationship. The created mapping relationship is stored in a database for subsequent determination of the correctness of the target attribution results.

[0077] In this invention, the multiple attribution algorithms include at least a device number attribution algorithm, a channel package attribution algorithm, and an IP and UA fuzzy attribution algorithm. The step of determining the target attribution result from the multiple initial attribution results according to priority rules includes: prioritizing the multiple attribution algorithms to obtain corresponding priority rules, wherein the device number attribution algorithm has the first priority, the IP and UA fuzzy attribution algorithm has the second priority, and the channel package attribution algorithm has the third priority; determining the initial attribution result corresponding to the highest priority attribution algorithm from the multiple initial attribution results according to the priority rules; and determining the initial attribution result corresponding to the highest priority attribution algorithm as the target attribution result for the device.

[0078] In embodiments of the present invention, such as Figure 2 As shown, the various attribution algorithms for channel attribution of devices in this invention include at least the device number attribution algorithm, the channel package attribution algorithm, and the IP and UA fuzzy attribution algorithm. The accuracy of these three attribution algorithms is determined through linear analysis. Based on their accuracy, the various attribution algorithms are prioritized to obtain corresponding priority rules. In these priority rules, the device number attribution algorithm has the highest priority (first priority), the IP and UA fuzzy attribution algorithm has the next lower priority (second priority), and the channel package attribution algorithm has the lowest priority (third priority). This yields the priority rules for these three attribution algorithms. After obtaining the priority rules, channel attribution of devices is performed using these three attribution algorithms. The initial attribution result corresponding to the highest priority attribution algorithm is determined, and this initial attribution result is set as the target attribution result for the device. If the device has an initial attribution result corresponding to the first-priority device number attribution algorithm, the initial attribution result corresponding to the device number attribution algorithm is determined as the target attribution result. If the device cannot be attributed using the device number attribution algorithm, it is determined whether it has an initial attribution result corresponding to the second-priority IP and UA fuzzy attribution algorithm. If it has an initial attribution result corresponding to the second-priority IP and UA fuzzy attribution algorithm, the initial attribution result is determined as the target attribution result for the device. If the device cannot be attributed using either the device number attribution algorithm or the IP and UA fuzzy attribution algorithm, the initial attribution result corresponding to the third-priority channel packet attribution algorithm is determined as the target attribution result for the device.

[0079] In this invention, determining the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rule includes: performing a union deduplication process on the initial attribution result determined by the device number attribution algorithm and the initial attribution result determined by the IP and UA fuzzy attribution algorithm to obtain a first deduplication result; performing a union deduplication process on the first deduplication result and the initial attribution result corresponding to the channel package attribution algorithm to determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results.

[0080] In an embodiment of the present invention, the initial attribution result determined by the device number attribution algorithm and the initial attribution result determined by the IP and UA fuzzy attribution algorithms of the device are first subjected to union deduplication to obtain a first deduplication result. This first deduplication result is: when the device has an initial attribution result corresponding to the device number attribution algorithm, the first deduplication result is the initial attribution result corresponding to the device number attribution algorithm; when the device does not have an initial attribution result corresponding to the device number attribution algorithm, the first deduplication result is the initial attribution result corresponding to the IP and UA fuzzy attribution algorithm. After obtaining the first deduplication result for the device, the first deduplication result is again subjected to union deduplication with the initial attribution result corresponding to the channel package attribution algorithm. When a first deduplication result is obtained, this first deduplication result is the initial attribution result corresponding to the highest priority attribution algorithm among the final multiple initial attribution results.

[0081] In this invention, the method further includes: calculating attribution precision and attribution recall based on whether the target attribution results of each device are correct, and obtaining attribution precision and attribution recall.

[0082] In embodiments of the present invention, attribution accuracy and attribution recall are calculated based on whether the target attribution results of each device are correct, thereby obtaining a more accurate evaluation method for the channel attribution accuracy and attribution recall of devices.

[0083] In this invention, the attribution accuracy is calculated based on whether the target attribution results of each device are correct. The attribution accuracy is obtained by: determining the number of correct target attribution results as a first parameter; determining the number of devices with account login as a second parameter; and dividing the first parameter by the second parameter to obtain the attribution accuracy.

[0084] In an embodiment of the present invention, the attribution accuracy is calculated as follows: the number of channels with correct target attribution results is determined as a first parameter, and the number of devices with account login is determined as a second parameter. The attribution accuracy of the channel attribution is obtained by dividing the first parameter by the second parameter.

[0085] For example, if there are 150 devices with target attribution results, 100 of which are logged in with an account, and 90 of these 100 devices have the target attribution results correctly judged, then the final channel attribution accuracy rate is 90 / 100 = 90%.

[0086] In this invention, attribution recall is calculated based on whether the target attribution result of each device is correct. The attribution recall is obtained by: determining the number of devices that log in within a time interval and have a mapping relationship with the account and have a target attribution result as a third parameter; determining the number of devices that download within a preset time range and have a login account within the time interval and have a mapping relationship with the account as a fourth parameter; and dividing the third parameter by the fourth parameter to obtain the attribution recall.

[0087] In an embodiment of the present invention, the attribution recall rate is calculated as follows: the number of devices that logged into an account within a time interval, have a mapping relationship stored in the database, and also have the target attribution result is determined as the third parameter. The number of devices that downloaded within a preset time range, logged into the account within the time interval, and have a mapping relationship stored in the database is determined as the fourth parameter. The attribution recall rate for channel attribution is obtained by dividing the third parameter by the fourth parameter.

[0088] The time interval can be set according to the actual application scenario, such as 24 hours, 48 ​​hours, etc., without specific limitations here. The preset time range can also be set according to the actual application scenario, such as 24 hours, 48 ​​hours, etc., without specific limitations here.

[0089] For example, if 300 new accounts are registered through internet channels, and 150 of these 300 accounts log in on 148 newly downloaded apps, and 120 of these 148 devices have the target attribution result, then the recall rate is: 120 / 148 = 80.1%.

[0090] In this invention, the methods for placing advertisements include at least two methods: mandatory registration and download.

[0091] In embodiments of the present invention, such as Figure 3 As shown, the advertising methods of this invention include strong registration advertising, in which users directly register an account by clicking on the H5 page. At this time, the registered user ID and the corresponding channel ID can be directly obtained, thereby establishing a mapping relationship between the two. The advertising methods of this invention also include download advertising, in which users directly download the channel package and APP by clicking on the advertisement, and then register through the APP, without having to be redirected to the activity page and the unified download page.

[0092] A second aspect of the present invention provides a multi-channel attribution system, such as... Figure 4 As shown, the system 400 includes:

[0093] The mapping relationship determination module 401 is used to obtain and store the mapping relationship between the registered user ID and the channel ID when the user registers an account through the device;

[0094] Attribution module 402 is used to attribute the device using multiple attribution algorithms to obtain multiple initial attribution results;

[0095] The target attribution result determination module 403 is used to determine the target attribution result from the plurality of initial attribution results according to priority rules;

[0096] The target attribution result accuracy determination module 404 is used to determine whether the target attribution result is correct based on the mapping relationship.

[0097] Optionally, the mapping relationship determination module includes:

[0098] The Channel ID tracking module is used to preset the channel ID of the advertising delivery channel through tracking logs;

[0099] The Channel ID and Registered User ID Acquisition Module is used to obtain the Channel ID and Registered User ID recorded in the event tracking log of an ad when a user registers an account by clicking on an ad through a device.

[0100] The mapping relationship determination submodule is used to create and store the corresponding mapping relationship based on the obtained channel ID and registered user ID.

[0101] Optionally, the multiple attribution algorithms include at least the device number attribution algorithm, the channel packet attribution algorithm, and the IP and UA fuzzy attribution algorithm;

[0102] The target attribution result determination module includes:

[0103] The priority sorting module is used to sort the various attribution algorithms by priority and obtain the corresponding priority rules. In the priority rules, the device number attribution algorithm is the first priority, the IP and UA fuzzy attribution algorithm is the second priority, and the channel package attribution algorithm is the third priority.

[0104] The first target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rules;

[0105] The second target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm as the target attribution result of the device.

[0106] Optionally, the system further includes:

[0107] The precision and recall determination module is used to calculate attribution precision and attribution recall based on whether the target attribution results of each device are correct, and obtain attribution precision and attribution recall.

[0108] A third aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the multi-channel attribution method described in the first aspect of the present invention.

[0109] A fourth aspect of the present invention provides a readable storage medium storing a program that, when executed by a processor, implements the steps of the multi-channel attribution method described in the first aspect of the present invention.

[0110] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0111] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0112] The foregoing has provided a detailed description of the multi-channel attribution method, system, electronic device, and readable storage medium provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A multi-channel attribution method, characterized in that, The method includes: When a user registers an account through a device, the mapping relationship between the registered user ID and the channel ID is obtained and stored; The device is attributed using multiple attribution algorithms to obtain multiple initial attribution results; Based on priority rules, a target attribution result is determined from the plurality of initial attribution results; Based on the mapping relationship, determine whether the target attribution result is correct; The various attribution algorithms include at least the device number attribution algorithm, the channel packet attribution algorithm, and the IP and UA fuzzy attribution algorithm; The step of determining the target attribution result from the plurality of initial attribution results according to the priority rule includes: The various attribution algorithms are prioritized to obtain corresponding priority rules. In the priority rules, the device number attribution algorithm is the first priority, the IP and UA fuzzy attribution algorithm is the second priority, and the channel package attribution algorithm is the third priority. Based on the priority rule, determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results; The initial attribution result corresponding to the highest priority attribution algorithm is determined as the target attribution result of the device. Specifically, the accuracy of the various attribution algorithms is determined through linear analysis, and the various attribution algorithms are prioritized according to their accuracy to obtain the priority rules.

2. The multi-channel attribution method according to claim 1, characterized in that, The step of obtaining and storing the mapping relationship between the registered user ID and the channel ID when a user registers an account through a device includes: The channel ID of the advertising delivery channel is preset through the event tracking logs; When a user registers an account by clicking on an ad through a device, obtain the channel ID and the registered user ID recorded in the ad's event logs; Based on the obtained channel ID and registered user ID, create and store the corresponding mapping relationship.

3. The multi-channel attribution method according to claim 1, characterized in that, The step of determining the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rule includes: The initial attribution result determined by the device number attribution algorithm and the initial attribution result determined by the IP and UA fuzzy attribution algorithm are combined to remove duplicates, and a first deduplication result is obtained. The first deduplication result is combined with the initial attribution result corresponding to the channel package attribution algorithm to perform deduplication, and the initial attribution result corresponding to the attribution algorithm with the highest priority among the multiple initial attribution results is determined.

4. The multi-channel attribution method according to claim 1, characterized in that, The step of determining whether the target attribution result is correct based on the mapping relationship includes: When the channel ID in the target attribution result is the same as the channel ID corresponding to the registered user ID in the mapping relationship, the target attribution result is determined to be correct. If the channel ID in the target attribution result is different from the channel ID corresponding to the registered user ID in the mapping relationship, the target attribution result is determined to be incorrect.

5. The multi-channel attribution method according to claim 1, characterized in that, The method further includes: Attribution precision and attribution recall are calculated based on whether the target attribution results of each device are correct.

6. The multi-channel attribution method according to claim 5, characterized in that, Attribution accuracy is calculated based on whether the target attribution results of each device are correct, and the attribution accuracy is obtained, including: The number of correct attributions for the target attribution result is determined as the first parameter; The number of devices with account logins is determined as the second parameter; The attribution accuracy is obtained by dividing the first parameter by the second parameter.

7. The multi-channel attribution method according to claim 5, characterized in that, Attribution recall is calculated based on whether the target attribution results for each device are correct, and the attribution recall rate is obtained, including: The number of devices that log in within a time interval, have a mapping relationship, and have a target attribution result is determined as the third parameter; The number of devices that download within a preset time range and log in to an account within that time range, and that account has a mapping relationship, is determined as the fourth parameter; The attribution recall rate is obtained by dividing the third parameter by the fourth parameter.

8. The multi-channel attribution method according to claim 2, characterized in that, The methods for placing advertisements include at least two types: mandatory registration and download-based advertising.

9. A multi-channel attribution system, characterized in that, The system includes: The mapping relationship determination module is used to obtain and store the mapping relationship between the registered user ID and the channel ID when a user registers an account through a device; The attribution module is used to attribute the device using multiple attribution algorithms to obtain multiple initial attribution results; The target attribution result determination module is used to determine the target attribution result from the plurality of initial attribution results according to priority rules; The target attribution result accuracy determination module is used to determine whether the target attribution result is correct based on the mapping relationship; The various attribution algorithms include at least the device number attribution algorithm, the channel packet attribution algorithm, and the IP and UA fuzzy attribution algorithm; The target attribution result determination module includes: The priority sorting module is used to sort the various attribution algorithms by priority and obtain the corresponding priority rules. In the priority rules, the device number attribution algorithm is the first priority, the IP and UA fuzzy attribution algorithm is the second priority, and the channel package attribution algorithm is the third priority. The first target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm from the plurality of initial attribution results according to the priority rules; The second target attribution result determination module is used to determine the initial attribution result corresponding to the highest priority attribution algorithm as the target attribution result of the device. Specifically, the accuracy of the various attribution algorithms is determined through linear analysis, and the various attribution algorithms are prioritized according to their accuracy to obtain the priority rules.

10. The multi-channel attribution system according to claim 9, characterized in that, The mapping relationship determination module includes: The Channel ID tracking module is used to preset the channel ID of the advertising delivery channel through tracking logs; The Channel ID and Registered User ID Acquisition Module is used to obtain the Channel ID and Registered User ID recorded in the event tracking log of an ad when a user registers an account by clicking on an ad through a device. The mapping relationship determination submodule is used to create and store the corresponding mapping relationship based on the obtained channel ID and registered user ID.

11. The multi-channel attribution system according to claim 9, characterized in that, The system also includes: The precision and recall determination module is used to calculate attribution precision and attribution recall based on whether the target attribution results of each device are correct, and obtain attribution precision and attribution recall.

12. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in memory, it implements the steps of a multi-channel attribution method as described in any one of claims 1-8.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements a multi-channel attribution method as described in any one of claims 1-8.

Images