Advertisement attribution method and device

Abstract

The invention discloses an advertisement attribution method and device. In the scheme, in response to a target application package installed by a user, installation information and an attribution strategy of the target application package are acquired; and matching target advertisement contact data from a target database based on the installation information. And configuring a target advertisement attribution model based on the attribution strategy. And inputting the target advertisement contact data into the target advertisement attribution model to obtain an advertisement attribution result. According to the scheme, the target database is utilized to associate the advertisement contact data in the time dimension, and the target advertisement attribution model flexibly configured based on the attribution strategy is combined, so that the customized configuration of the advertisement attribution model is realized, and the accuracy of advertisement attribution is remarkably improved in the face of complex and changeable advertisement putting scenes.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an advertising attribution method and device. Background Art

[0002] In the current digital advertising environment, the challenges facing advertisers are becoming increasingly complex and varied. With the rapid development of Internet technology and the popularization of smart devices, users are no longer limited to receiving information through a single channel or device, but are frequently interacting between multiple touchpoints (such as search engines, social media and video platforms) and multiple devices (such as mobile devices, tablets and personal computers). This cross-channel and cross-device interaction model not only greatly enriches the user's online experience, but also brings unprecedented challenges to the accurate attribution of advertising effects.

[0003] Advertising attribution models are key tools for evaluating the effectiveness of advertising campaigns. They analyze and track user behavior after exposure to ads to determine which ad touchpoints or interactions on devices ultimately led to user conversions (such as purchases, downloads, or registrations). However, most existing advertising attribution models are designed based on fixed and single rules, such as last-click attribution and linear attribution models. Faced with the increasingly fragmented user behavior and highly complex interaction scenarios, they are unable to accurately and effectively capture and evaluate the true effects of advertising at each touchpoint, resulting in low attribution accuracy. Summary of the invention

[0004] Based on the above problems, the present application provides an advertising attribution method and device, the purpose of which is to perform advertising attribution more accurately when facing complex interactive scenarios.

[0005] The embodiments of the present application disclose the following technical solutions:

[0006] The first aspect of the present application provides an advertising attribution method, the method comprising:

[0007] In response to the user installing the target application package, obtaining installation information and attribution strategy of the target application package; the installation information includes time information and device information, the device information includes multiple device identifiers, and the multiple device identifiers indicate the same device; the attribution strategy includes a filtering strategy and a processing strategy;

[0008] Matching target advertising contact data from a target database based on the installation information; the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices; the time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index; wherein each advertising index corresponds to one piece of advertising contact data; the target advertising contact data contains a device identifier;

[0009] Configure a target advertising attribution model based on the attribution strategy;

[0010] The target advertising contact data is input into the target advertising attribution model to obtain an advertising attribution result; the advertising attribution result includes at least one piece of advertising contact data.

[0011] In an optional implementation, configuring a target advertisement attribution model based on the attribution strategy includes:

[0012] Generate configuration rules based on the attribution strategy; the configuration rules include filtering rules and processing rules;

[0013] Building an attribution execution process based on the configuration rules;

[0014] The attribution execution process is configured into a target advertising attribution model.

[0015] In an optional implementation, the method for constructing the target database includes:

[0016] Receiving a plurality of pieces of advertising contact data transmitted by a plurality of advertising platforms; the advertising contact data including a plurality of device identifiers;

[0017] Based on the multiple pieces of advertising contact data, generate a time index sequence for each device identifier in each piece of advertising contact data;

[0018] The time index sequence of each device identification in each piece of advertising contact data is stored in the target database.

[0019] In an optional implementation, matching target advertisement contact data from a target database based on the installation information includes:

[0020] Based on the device information, a first time index sequence is screened out from the target database; the first time index sequence is a time index sequence of multiple device identifiers corresponding to the device information;

[0021] Based on the time information, respectively generate matching time windows for the multiple device identifiers;

[0022] Target advertising contact data is matched from the first time index sequence based on matching time windows of the multiple device identifiers.

[0023] In an optional implementation, the advertising attribution method also includes:

[0024] In response to a user opening a target application for the first time after installing the target application package, a unique identifier is generated and stored in the target application package; the unique identifier uniquely indicates a device on which the user installs the target application package; the device corresponds to multiple device identifiers;

[0025] verifying whether the device is an abnormal device based on the unique identifier;

[0026] If the device is an abnormal device, multiple device identifiers corresponding to the device are marked as abnormal device identifiers.

[0027] In an optional implementation, the verifying whether the device is an abnormal device based on the unique identifier includes:

[0028] If the number of changes of the device identification corresponding to the unique identifier reaches a threshold, the device is an abnormal device;

[0029] If the user behavior of the device identification corresponding to the unique identifier is abnormal, the device is an abnormal device.

[0030] In an optional implementation, the advertising attribution method also includes:

[0031] The time index sequence of the abnormal device identification in the target database is disabled.

[0032] In an optional implementation, each of the multiple device identifiers has a corresponding identification bit; after matching the target advertising contact data from the target database based on the installation information, the method further includes:

[0033] According to the identification bit sequence of the multiple device identifications in the device information, the device identifications included in the target advertisement contact data are compared one by one to obtain a comparison result;

[0034] A credibility score for the target advertising contact data is generated based on the comparison result; the credibility score indicates the attribution credibility of the target advertising contact data.

[0035] A second aspect of the present application provides an advertising attribution device, the device comprising:

[0036] an acquisition module, configured to acquire, in response to a user installing a target application package, installation information and an attribution strategy of the target application package; the installation information includes time information and device information, the device information includes multiple device identifiers, and the multiple device identifiers indicate the same device; the attribution strategy includes a filtering strategy and a processing strategy;

[0037] A matching module, configured to match target advertising contact data from a target database based on the installation information; the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices; the time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index; wherein each advertising index corresponds to one piece of advertising contact data; the target advertising contact data contains a device identifier;

[0038] A configuration module, used to configure a target advertisement attribution model based on the attribution strategy;

[0039] The attribution module is used to input the target advertising contact data into the target advertising attribution model to obtain an advertising attribution result; the advertising attribution result includes at least one piece of advertising contact data.

[0040] In an optional implementation, the configuration module includes:

[0041] A configuration rule generating unit, configured to generate configuration rules based on the attribution strategy; the configuration rules include filtering rules and processing rules;

[0042] An execution process building unit, used to build an attribution execution process based on the configuration rules;

[0043] A model configuration unit is used to configure the attribution execution process into a target advertising attribution model.

[0044] Compared with the prior art, this application has the following beneficial effects:

[0045] In the technical solution of the present application, first, in response to the user installing the target application package, the installation information and attribution strategy of the target application package are obtained; then, the target advertising contact data is matched from the target database based on the installation information; then, the target advertising attribution model is configured based on the attribution strategy; finally, the target advertising contact data is input into the target advertising attribution model to obtain the advertising attribution result. The technical solution of the present application establishes the association of advertising contact data in the time dimension through the target database, so that the advertising attribution method can effectively cope with complex and changeable multi-touch and multi-device interaction scenarios, and significantly improve the accuracy of advertising attribution. By adopting a target advertising attribution model that is flexibly configured based on the attribution strategy, the customization of the attribution model for different scenarios is realized, so that when faced with complex and diverse advertising delivery scenarios, the accuracy of advertising attribution can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0047] Figure 1 A flow chart of an advertising attribution method provided in an embodiment of the present application;

[0048] Figure 2 A schematic diagram of a scenario for configuring an attribution strategy combination provided in an embodiment of the present application;

[0049] Figure 3 A flowchart of another advertising attribution method provided in an embodiment of the present application;

[0050] Figure 4 A schematic diagram of a scenario for matching target advertisement contact data provided in an embodiment of the present application;

[0051] Figure 5 A schematic diagram of an attribution execution process provided in an embodiment of the present application;

[0052] Figure 6 A schematic diagram of the structure of an advertising attribution device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] As described above, current advertising attribution methods are unable to perform accurate and effective advertising attribution when faced with complex interactive scenarios.

[0054] After research, the inventor proposed an advertising attribution method and device.

[0055] First, in response to the user installing the target application package, the installation information and attribution strategy of the target application package are obtained; then, the target advertising contact data is matched from the target database based on the installation information; then, the target advertising attribution model is configured based on the attribution strategy; finally, the target advertising contact data is input into the target advertising attribution model to obtain the advertising attribution result.

[0056] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] Advertising attribution refers to determining the channel attribution of an advertisement after it is exposed and clicked on multiple platforms or resources, resulting in a conversion (such as a user installing a target app package).

[0058] It should be noted that the user-related information involved in this application (including but not limited to device information, time information, etc.) are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0059] See also Figure 1 , which is a flow chart of an advertising attribution method provided in an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0060] S101 . In response to a user installing a target application package, obtaining installation information and an attribution policy of the target application package.

[0061] An application package is a collection of files that organizes and packages an application and all its related components and resources. It contains all the code, resource files (such as images, audio, etc.), configuration files, and metadata required to run the application.

[0062] In the embodiment of the present application, the target application package refers to the application package to which advertising attribution is to be performed.

[0063] The advertising attribution method provided in the embodiment of the present application responds to the user's operation of installing the target application package. First, the installation information of the target application package is obtained, and the installation information includes but is not limited to time information and device information.

[0064] The time information indicates the installation time of the target application package. For example, if the time information is 2024.07.25-12:00, it indicates that the target application package is installed at 12:00 on July 25, 2024.

[0065] The device information indicates the device on which the user installs the target application package. The device information includes multiple device identifiers. The multiple device identifiers indicate the same device, that is, the device on which the target application package is installed.

[0066] Device identification refers to an identifier that names a device at the hardware or software level, which is used to distinguish and identify different devices. The same device corresponds to multiple device identifications, such as the Identifier For Advertising (IDFA), the Open Anonymous Device Identifier (OAID), the International Mobile Equipment Identity (IMEI), and the Internet Protocol (IP) address.

[0067] Multiple device identifiers together constitute the uniqueness of the device, helping to avoid duplicate or incorrect attribution.

[0068] In the embodiment of the present application, while obtaining the installation information of the target application package, the attribution strategy of the target application package is obtained. The attribution strategy refers to the rules or methods used to determine which advertising contact data leads to the user installing the target application.

[0069] In the embodiment of the present application, the attribution strategy includes a filtering strategy and a processing strategy. The filtering strategy is used to filter and exclude advertising contact data that does not meet the conditions, and the processing strategy is responsible for determining the contribution of the filtered advertising contact data.

[0070] In an embodiment of the present application, the attribution strategy can be obtained from a preset configuration center, which includes a plurality of pre-written filtering strategies and a plurality of processing strategies. The appropriate strategy combination can be flexibly selected according to the characteristics of the target application package, the goals of the advertising campaign, and the characteristics of the user behavior data.

[0071] Figure 2 A schematic diagram of a scenario for configuring an attribution strategy combination provided in an embodiment of the present application. Figure 2 As shown, the configuration center includes pre-written attribution strategies, multiple filtering strategies, and multiple processing strategies. First, the corresponding attribution strategy is obtained from the configuration center, and then a processing flow combination is generated based on the obtained attribution strategy. Finally, the corresponding filtering strategy and processing strategy are obtained from the configuration center according to the processing flow combination, and combined into an attribution processing flow.

[0072] S102: Match target advertisement contact data from a target database based on the installation information.

[0073] In the technical solution of the present application, a target database is creatively constructed. The target database includes a time index sequence of multiple device identifiers corresponding to multiple devices. The time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index. Each advertising index corresponds to an advertising contact data.

[0074] The following describes how to construct the target database.

[0075] In order to build the target database, firstly, multiple advertising touch point data transmitted by multiple advertising platforms are received in real time.

[0076] In the embodiment of the present application, when an advertiser places an advertisement, the advertiser may place the same advertisement on multiple advertising platforms. According to the advertisement data specification of each platform, a corresponding network interface may be provided for each platform to receive the advertisement contact data transmitted by each advertising platform.

[0077] Ad touchpoint data refers to the data generated when a user sees, clicks, interacts with, or converts an ad. These data points reflect the user's response and behavior patterns to the ad. Ad touchpoint data includes multiple device identifiers.

[0078] Optionally, to facilitate subsequent processing and storage, after receiving the advertising contact data transmitted by multiple advertising platforms, the different advertising contact data are processed into advertising contact data with a unified standard.

[0079] Afterwards, based on the received multiple pieces of advertising contact data, a time index sequence of each device identification in each piece of advertising contact data is generated. The time index sequence of the device identification represents a series of advertising behaviors that occur over time under the device identification.

[0080] Finally, the time index sequence of each device identification in each piece of advertising contact data is stored in the target database.

[0081] In an example implementation, the ZSet structure of Redis can be used to store the time index sequence of device identifiers. Specifically, the event timestamp of the advertising contact data can be used as the score of the ZSet member, and the advertising index of the advertising contact data can be used as the member value, and then the parsed data can be stored in Redis through the ZADD command of ZSet.

[0082] When you need to query the advertising touchpoint data within a certain time window later, you can use the Redis ZRANGEBYSCORE command to obtain all advertising touchpoint data within the time window by specifying the timestamp value range (that is, the score value range).

[0083] In the embodiment of the present application, since the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices of multiple users transmitted by multiple advertising platforms, it is necessary to match the advertising contact data related to this installation behavior from the target database based on the installation information, that is, the target advertising contact data, for subsequent advertising attribution. The target advertising contact data contains the device identifier.

[0084] In the embodiment of the present application, the target advertising contact data is all advertising behaviors within a period of time before the user installs the target application package. The target advertising contact data includes multiple pieces of advertising contact data.

[0085] S103. Configure a target advertising attribution model based on the attribution strategy.

[0086] In an embodiment of the present application, the advertising attribution model is an attribution model with customizable rules, which can be flexibly configured based on the attribution strategy obtained in the aforementioned steps to obtain a target advertising attribution model.

[0087] S104: Input the target advertising contact data into the target advertising attribution model to obtain the advertising attribution result.

[0088] In an embodiment of the present application, the multiple advertising contact data included in the target advertising contact data matched in S102 are used as input to the target advertising attribution model. The target advertising attribution model processes the input multiple advertising contact data according to preset attribution strategies and parameters, calculates the contribution of each target advertising contact data to user behavior, and finally outputs the advertising attribution result.

[0089] In an embodiment of the present application, the advertising attribution result includes at least one piece of advertising contact data, that is, the advertising attribution result can be one piece of advertising contact data (attributing the current installation behavior to one piece of advertising contact data) or multiple pieces of advertising contact data (attributing the current installation behavior to multiple pieces of advertising contact data).

[0090] Optionally, the advertising attribution result also includes advertising effectiveness data and the advertising platform that most effectively guides users to install the target application package. For example, the user may be exposed to and install the target application package through social media advertising, search engine advertising, video advertising, or other forms of online advertising.

[0091] The embodiment of the present application innovatively constructs a target database and successfully establishes the association of advertising touch point data in the time dimension, so that the advertising attribution method can effectively cope with complex and changeable multi-touch point and multi-device interaction scenarios, and significantly improve the accuracy of advertising attribution. In addition, by adopting a target advertising attribution model with flexible configuration based on attribution strategy, a targeted processing flow is constructed for different business scenarios, and the customization of the attribution model is realized. At the same time, the advertising attribution results output by the model can output one or more advertising touch point data according to different configurations, so that when facing complex and diverse advertising delivery scenarios, the accuracy of advertising attribution for different scenarios can be effectively improved.

[0092] Optionally, after obtaining the attribution results, the advertising platform is called back according to the attribution results to optimize the advertising model. A crowd portrait is constructed based on the attribution results, which depicts key features such as user interest preferences, consumption habits, and active time periods. Based on the crowd portrait, users from specific advertising sources are accurately pushed to the client, and personalized operation activities are customized for them.

[0093] Figure 3 Another flow chart of an advertising attribution method provided in an embodiment of the present application. In the embodiment introduced by this figure, a more detailed description is provided for the implementation of the advertising attribution method.

[0094] like Figure 3 The ad attribution methodology shown includes the following steps:

[0095] S301. In response to a user installing a target application package, obtaining installation information and attribution policy of the target application package.

[0096] The implementation of S301 is basically the same as that of S101 in the method embodiment described above, and will not be described in detail here. For related technical implementations, please refer to the description of S101 described above.

[0097] S302: Filter out a first time index sequence from a target database based on device information.

[0098] In the embodiment of the present application, the first time index sequence is a time index sequence of multiple device identifiers corresponding to the device information.

[0099] In an exemplary implementation, the multiple device identifications included in the device information are device identification A and device identification B, and the time index sequence of device identification A and device identification B is filtered out from the target database as the first time index sequence.

[0100] S303: Generate matching time windows for multiple device identifiers based on the time information.

[0101] In the embodiment of the present application, the matching time window refers to the time period of valid advertising contact data of a certain device identification. The advertising contact data generated during this time period is considered to be possibly related to the user's installation behavior.

[0102] In an embodiment of the present application, different device identifiers correspond to different matching time windows due to their different stability and reliability. For example, OAID is provided by the device manufacturer or operating system, which is relatively stable, reliable and not easy to change. Therefore, a longer matching time window, such as 24 hours, can be set for it. This means that all advertising contact data occurring within 24 hours before the user installs the target application package are taken into account. The IP address may change due to the user changing the network environment (such as switching from a home network to a corporate network), so its stability is poor. In order to reduce the risk of misattribution, a shorter matching time window, such as 3 hours, is set for it. This means that only the advertising contact data occurring through the same IP address within 3 hours before the user installs the target application package will be considered.

[0103] In the embodiment of the present application, the corresponding matching time window is generated according to the time information of the user installing the target application package (such as "2024.07.25-12:00") and the length of the matching time window corresponding to each device identifier (such as 24 hours for OAID and 3 hours for IP). At the same time, the shortest time required for the user to download and install the target application package must also be considered.

[0104] In an example implementation, if the shortest time required for a user to download and install the target application package is 5 minutes, then for OAID, the matching time window is "2024.07.24-11:55" to "2024.07.25-11:55". For IP address, the matching time window is "2024.07.25-08:55" to "2024.07.25-11:55".

[0105] S304: Match target advertisement contact data from the first time index sequence based on the matching time windows of the multiple device identifiers.

[0106] After the matching time windows are generated, target advertisement contact data that meets the conditions is matched from the first time index sequence based on the matching time windows.

[0107] Figure 4 A schematic diagram of a scenario for matching target advertising contact data provided by an embodiment of the present application. Figure 4 As shown, the first time index sequence includes the time index sequence of device identification A and device identification B. The matching time window A of device identification A includes timestamp 2 to timestamp 4, and the matching time window B of device identification B includes timestamp 3 to timestamp 4. The target advertising contact data matched according to the matching windows include advertising contact data corresponding to advertising index 2 (including device identification A), advertising index 3 (including device identification A and device identification B), and advertising index 4 (including device identification A and device identification B).

[0108] S305: Generate configuration rules based on the attribution strategy.

[0109] In the embodiment of the present application, the configuration rules include filtering rules and processing rules.

[0110] According to the attribution strategy, the system generates a series of configuration rules, including filtering rules and processing rules. Filtering rules are used to remove advertising touch point data that does not meet the attribution requirements, such as repeated clicks, invalid exposures, etc. Processing rules define how to process the remaining advertising touch point data, such as priority sorting, weight allocation, etc.

[0111] In an example implementation, the configuration rules may be organized into a list, each element in the list is a Map, and the Map contains two key-value pairs: key and value.

[0112] In the embodiments of the present application, the key is used as an identifier of a specific rule (such as "filter" or "attribute"), and the value is a list or another Map containing specific rules or conditions.

[0113] An example of the structure of a configuration rule list is as follows:

[0114]

[0115] In the above examples, the "TODO" field is used as a link to the next node or as a placeholder to mark the MCA node.

[0116] In an example implementation, the configuration rule list is as follows:

[0117]

[0118]

[0119]

[0120]

[0121] S306: Construct an attribution execution process based on the configuration rules.

[0122] In the embodiment of the present application, a corresponding attribution execution process is constructed based on the configuration rules generated in the previous step. The attribution execution process is a linked list structure, in which each node contains specific filtering rules or processing rules. The last node of the attribution execution process is the Main Conversion Attribution (MCA) node, which is used to output the final attribution result.

[0123] In the embodiment of the present application, the Key and Value in each Map in the configuration rule list can be parsed by traversing the configuration rule list, wherein the Key determines the type of the node (filter node or processing node), and the value contains specific rules or conditions.

[0124] According to the parsing results, a corresponding process node is created for each Map. The nodes are linked in sequence to form a linked list structure. Each node contains specific rules for execution (filtering rules or processing rules).

[0125] The last node of the linked list structure is set as the MCA node, which outputs the final attribution result.

[0126] Figure 5 A schematic diagram of an attribution execution process provided in an embodiment of the present application. Figure 5As shown, the attribution execution process includes that after the advertising contact data is filtered by filtering rules 1 and filtering rules 2 in the filter node, it points to the judgment node through TODO. The judgment node determines whether the filtered advertising contact data meets judgment condition 1. If judgment condition 1 is met, it jumps to the attribute node. After the filtered advertising contact data is processed according to processing rules 1 and processing rules 2 in the attribute node, it points to the judgment node through TODO. It determines whether the processed advertising contact data meets judgment condition 2. If judgment condition 2 is met, it jumps to the MCA node and outputs the attribution result.

[0127] S307: Configure the attribution execution process into the target advertisement attribution model.

[0128] In an embodiment of the present application, the constructed attribution execution process is configured into the target advertising attribution model, so that the model can filter and process the advertising contact data according to the defined filtering rules and processing rules, and output the attribution results.

[0129] S308: Input the target advertising contact data into the target advertising attribution model to obtain the advertising attribution result.

[0130] The implementation of S308 is basically the same as that of S104 in the method embodiment described above, and will not be described in detail here. For related technical implementations, please refer to the description of S104 described above.

[0131] Optionally, in order to avoid the influence of advertising cheating on the advertising attribution results and improve the accuracy of advertising attribution, the above-mentioned advertising attribution method further includes:

[0132] Step 1: In response to the user opening the target application for the first time after installing the target application package, a unique identifier (eg, UUID) is generated and stored in the locally installed target application package.

[0133] The unique identifier uniquely indicates the device on which the user installs the target application package. The device corresponds to multiple device identifiers (such as IDFA, OAID, IMEI, and IP address, etc.).

[0134] Step 2: Verify whether the device is an abnormal device based on the unique identifier in the target application package.

[0135] In the embodiment of the present application, if the number of changes in the device identification corresponding to the unique identifier reaches a threshold, the device is an abnormal device. The threshold can be set according to actual conditions, such as 3 times or 5 times, which is not specifically limited here.

[0136] In an example implementation, the unique identifier is a UUID, and the multiple device identifiers corresponding to the device are IDFA, OAID, and IMEI, then the UUID corresponds to the three device identifiers IDFA, OAID, and IMEI. If the OAID corresponding to the UUID changes three times, the device is an abnormal device.

[0137] In the embodiment of the present application, if the user behavior of the device identification corresponding to the unique identifier is abnormal, the device is an abnormal device.

[0138] In an example implementation, when the same device identifier repeatedly registers a large number of application accounts corresponding to the target application packages, it is determined that the user behavior of the device identifier is abnormal; when the number of times the application account corresponding to the same device identifier receives a reward exceeds a preset threshold, it is determined that the user behavior of the device identifier is abnormal, wherein the reward is a reward from the application corresponding to the target application package to the newly registered application account.

[0139] Step 3: If the device is an abnormal device, multiple device identifiers corresponding to the device are marked as abnormal device identifiers to prevent these abnormal device identifiers from being misused in the future, thereby affecting the accuracy of advertising attribution.

[0140] In an exemplary implementation, for an abnormal device like the one in the aforementioned example, its corresponding IDFA, OAID and IMEI are all marked as abnormal device identifiers.

[0141] In the embodiment of the present application, each device is accurately identified by using a unique identifier stored in the local installation package, and the unique identifier is used to verify abnormal devices. Accurate identification of advertising fraud scenarios is achieved, effectively avoiding the problem of reduced accuracy in identifying advertising fraud scenarios due to limited device information collection (such as privacy policy restrictions), thereby improving the accuracy of advertising attribution.

[0142] In order to further avoid the impact of advertising fraud on advertising attribution results, different abnormal device identification criteria can be set according to actual needs.

[0143] In an example implementation, if a device identifier that is not an IP address generates a large amount (e.g., 500) of advertising contact data within a short time window (e.g., 1 hour), the device identifier is determined to be an abnormal device identifier; if a device identifier has advertising contact data associated with multiple different IP addresses (e.g., more than 8) within a longer time window (e.g., 24 hours), the device identifier is determined to be an abnormal device identifier; if a unique hardware identifier (e.g., IMEI) of a device frequently appears in a short period of time with multiple different combinations of other device identifiers (e.g., OAID and MAC address), and the number exceeds a certain threshold (e.g., more than 5), the corresponding device identifier of the device is determined to be an abnormal device identifier; if a device identifier continues to generate advertising contact data on the same advertising platform over a long period of time (e.g., 9 consecutive months), but is never accompanied by download or installation behavior, the device identifier is determined to be an abnormal device identifier. The above criteria for determining abnormal device identifiers are only exemplary descriptions. In actual applications, the criteria for determining abnormal device identifiers can be set according to actual conditions and are not specifically limited here.

[0144] In an optional implementation, the time index sequence of the abnormal device identification in the target database is disabled.

[0145] In an embodiment of the present application, after disabling the time index sequence of the abnormal device identification in the target database, any advertising activity data or user behavior data related to the abnormal device identification will be ignored during advertising attribution, thereby avoiding affecting the advertising attribution results.

[0146] In an optional implementation, each of the multiple device identifiers has a corresponding identifier. In order to ensure the accuracy of advertising attribution when the device information is reduced due to privacy policy restrictions, after the aforementioned step S102, the advertising attribution method further includes:

[0147] Step 1: According to the identification bit sequence of multiple device identifications in the device information, the device identifications contained in the target advertising contact data are compared one by one to obtain a comparison result.

[0148] In the embodiment of the present application, the comparison result of a target advertising contact data is the number of times the device identifier in the target advertising contact data is the same as the number of multiple device identifiers in the device information. For example, the multiple device identifiers in the device information are device identifier A, device identifier B, device identifier C, device identifier D, and device identifier E in the order of the symbols, and the device identifiers in a certain target advertising contact data are device identifier A, device identifier D, and device identifier E. After comparison, the comparison result of the target advertising contact data is 3.

[0149] Step 2: Generate a credibility score for the target advertising touch point data based on the comparison result; the credibility score indicates the attribution credibility of the target advertising touch point data.

[0150] In the embodiment of the present application, the credibility score is the ratio of the number of device identifiers in the target advertising contact data that are the same as the multiple device identifiers in the device information to the total number of device identifiers included in the device information. For example, the credibility score of the target advertising contact data in the above example is 3 / 5, i.e. 60%.

[0151] In the embodiment of the present application, by generating a corresponding credibility score for each target advertising contact data, the credibility score threshold in the attribution strategy can be flexibly set to screen effective advertising contact data according to different requirements for the accuracy of the attribution results in different delivery scenarios. For example, in a scenario where high-precision attribution results are required, a higher credibility score threshold (such as more than 80%) can be set to ensure that the selected advertising contact data has a higher credibility; in a scenario where the accuracy of the attribution results is not high, a lower credibility score threshold (such as more than 50%) can be set to expand the selection range of advertising contact data, thereby potentially obtaining richer advertising effect data.

[0152] Based on the advertising attribution method provided in the aforementioned embodiment, the present application also provides an advertising attribution device accordingly. Figure 6 This is a schematic diagram of the structure of an advertising attribution device provided in an embodiment of the present application. Figure 6 As shown, the advertisement attribution device includes: an acquisition module 601, a matching module 602, a configuration module 603 and an attribution module 604.

[0153] The acquisition module 601 is used to acquire the installation information and attribution policy of the target application package in response to the user installing the target application package.

[0154] In an embodiment of the present application, the installation information includes time information and device information, the device information includes multiple device identifiers, and the multiple device identifiers indicate the same device.

[0155] Attribution strategies include filtering strategies and processing strategies.

[0156] The matching module 602 is used to match the target advertisement contact data from the target database based on the installation information acquired by the acquisition module 601 .

[0157] In an embodiment of the present application, the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices. The time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index. Each advertising index corresponds to an advertising contact data, and the target advertising contact data contains a device identifier.

[0158] In an embodiment of the present application, the target advertisement contact data includes multiple advertisement contact data.

[0159] Configuration module 603, configured to configure a target advertisement attribution model based on the attribution strategy acquired by acquisition module 601;

[0160] The attribution module 604 is used to input the multiple pieces of advertising contact data included in the target advertising contact data matched by the matching module 602 into the target advertising attribution model configured by the configuration module 603 to obtain the advertising attribution result.

[0161] In an embodiment of the present application, the advertising attribution result includes at least one piece of advertising contact data, that is, the advertising attribution result can be one piece of advertising contact data (attributing the current installation behavior to one piece of advertising contact data) or multiple pieces of advertising contact data (attributing the current installation behavior to multiple pieces of advertising contact data).

[0162] The embodiment of the present application combines the functions of the acquisition module 601, the matching module 602, the configuration module 603 and the attribution module 604 to achieve flexible configuration of the target advertising attribution model based on different attribution strategies, and the advertising attribution result output by the model can output one or more advertising contact data according to different configurations. When faced with complex and diverse advertising delivery scenarios, it is possible to configure corresponding attribution models for different scenarios and output different advertising attribution results according to the corresponding configurations, thereby improving the accuracy of advertising attribution in complex and changeable advertising delivery scenarios.

[0163] In an optional implementation, the configuration module 603 includes:

[0164] The configuration rule generation unit is used to generate configuration rules based on the attribution strategy. The configuration rules include filtering rules and processing rules.

[0165] The execution process building unit is used to build an attribution execution process based on the configuration rules generated by the configuration rule generating unit.

[0166] The model configuration unit is used to configure the attribution execution process built by the execution process building unit into the target advertising attribution model.

[0167] In an optional implementation, the matching module 602 is specifically configured to:

[0168] Based on the device information, a first time index sequence is screened out from the target database. The first time index sequence is a time index sequence of multiple device identifiers corresponding to the device information.

[0169] Based on the time information, matching time windows of multiple device identifiers are generated respectively.

[0170] Target advertising contact data is matched from the first time index sequence based on a matching time window of multiple device identifiers.

[0171] In an optional implementation, the advertisement attribution device further includes a target database construction module, which is used to:

[0172] Receive multiple pieces of advertising touch point data transmitted by multiple advertising platforms. The advertising touch point data includes multiple device identifiers.

[0173] Based on multiple pieces of advertising contact data, a time index sequence for each device identifier in each piece of advertising contact data is generated.

[0174] The time index sequence of each device identifier in each piece of advertising contact data is stored in the target database.

[0175] In an optional implementation, the advertising attribution apparatus further includes a unique identification generation module, an abnormal device verification module, and an abnormal device identification marking module.

[0176] The unique identification generating module is used to generate a unique identifier and store it in the target application package in response to the user opening the target application for the first time after installing the target application package.

[0177] In the embodiment of the present application, the unique identifier uniquely indicates the device on which the user installs the target application package. The device corresponds to multiple device identifiers.

[0178] The abnormal device verification module is used to verify whether the device is an abnormal device based on the unique identifier generated by the unique identifier generation module.

[0179] In an embodiment of the present application, if the number of changes in the device identification corresponding to the unique identifier reaches a threshold, the abnormal device verification module verifies that the device is an abnormal device; if the user behavior of the device identification corresponding to the unique identifier is abnormal, the abnormal device verification module verifies that the device is an abnormal device.

[0180] The abnormal device identification marking module is used to mark multiple device identifications corresponding to the device as abnormal device identifications if the abnormal device verification module verifies that the device is an abnormal device.

[0181] In an optional implementation, the advertising attribution apparatus further includes a time index sequence disabling module, which is used to disable the time index sequence of abnormal device identification in the target database.

[0182] In an optional implementation, each of the multiple device identifiers has a corresponding identification bit, and the advertising attribution device further includes a trust score generation module, which is used to compare the identification bits of the multiple device identifiers in the device information with the device identifiers included in the target advertising contact data one by one to obtain a comparison result. A trust score for the target advertising contact data is generated based on the comparison result. The trust score indicates the attribution credibility of the target advertising contact data.

[0183] The embodiment of the present application comprehensively expands and optimizes the advertising attribution device by introducing a target database construction module, a unique identification generation module, an abnormal device verification module, an abnormal device identification marking module, a time index sequence disabling module, and a trust score generation module. The joint action of these modules establishes a close connection between advertising contact data in the time dimension, so that the advertising attribution device can efficiently cope with complex and changeable multi-touch and multi-device interaction scenarios, and significantly improve the accuracy of advertising attribution. At the same time, it effectively avoids the interference of advertising cheating on the advertising attribution results, enhances the flexibility of advertising attribution configuration, and thus improves the accuracy of the advertising attribution device as a whole when dealing with complex and diverse advertising delivery scenarios.

[0184] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device embodiment described above is merely schematic, in which the unit described as a separate component may or may not be physically separated, and the component prompted as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0185] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An advertising attribution method, characterized in that: The method comprises: In response to the user installing the target application package, obtaining installation information and attribution strategy of the target application package; the installation information includes time information and device information, the device information includes multiple device identifiers, and the multiple device identifiers indicate the same device; the attribution strategy includes a filtering strategy and a processing strategy; Matching target advertising contact data from a target database based on the installation information; the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices; the time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index; wherein each advertising index corresponds to one piece of advertising contact data; the target advertising contact data contains a device identifier; Configure a target advertising attribution model based on the attribution strategy; The target advertising contact data is input into the target advertising attribution model to obtain an advertising attribution result; the advertising attribution result includes at least one piece of advertising contact data.

2. The method according to claim 1, characterized in that The configuring the target advertisement attribution model based on the attribution strategy includes: Generate configuration rules based on the attribution strategy; the configuration rules include filtering rules and processing rules; Building an attribution execution process based on the configuration rules; The attribution execution process is configured into a target advertising attribution model.

3. The method according to claim 1, characterized in that The method for constructing the target database comprises: Receiving a plurality of pieces of advertising contact data transmitted by a plurality of advertising platforms; the advertising contact data including a plurality of device identifiers; Based on the multiple pieces of advertising contact data, generate a time index sequence for each device identifier in each piece of advertising contact data; The time index sequence of each device identification in each piece of advertising contact data is stored in the target database.

4. The method according to claim 3, characterized in that The step of matching target advertisement contact data from a target database based on the installation information includes: Based on the device information, a first time index sequence is screened out from the target database; the first time index sequence is a time index sequence of multiple device identifiers corresponding to the device information; Based on the time information, respectively generate matching time windows for the multiple device identifiers; Target advertising contact data is matched from the first time index sequence based on matching time windows of the multiple device identifiers.

5. The method according to claim 1, characterized in that The method further comprises: In response to a user opening a target application for the first time after installing the target application package, a unique identifier is generated and stored in the target application package; the unique identifier uniquely indicates a device on which the user installs the target application package; the device corresponds to multiple device identifiers; verifying whether the device is an abnormal device based on the unique identifier; If the device is an abnormal device, multiple device identifiers corresponding to the device are marked as abnormal device identifiers.

6. The method according to claim 5, characterized in that The verifying whether the device is an abnormal device based on the unique identifier includes: If the number of changes of the device identification corresponding to the unique identifier reaches a threshold, the device is an abnormal device; If the user behavior of the device identification corresponding to the unique identifier is abnormal, the device is an abnormal device.

7. The method according to claim 6, characterized in that The method further comprises: The time index sequence of the abnormal device identification in the target database is disabled.

8. The method according to claim 1, characterized in that The multiple device identifiers each have a corresponding identification bit; after matching the target advertising contact data from the target database based on the installation information, the method further includes: According to the identification bit sequence of the multiple device identifications in the device information, the device identifications included in the target advertisement contact data are compared one by one to obtain a comparison result; A credibility score for the target advertising contact data is generated based on the comparison result; the credibility score indicates the attribution credibility of the target advertising contact data.

9. An advertising attribution device, characterized in that: The device comprises: an acquisition module, configured to acquire, in response to a user installing a target application package, installation information and an attribution strategy of the target application package; the installation information includes time information and device information, the device information includes multiple device identifiers, and the multiple device identifiers indicate the same device; the attribution strategy includes a filtering strategy and a processing strategy; A matching module, configured to match target advertising contact data from a target database based on the installation information; the target database includes a time index sequence of multiple device identifiers corresponding to multiple devices; the time index sequence includes multiple advertising indexes and timestamps corresponding to each advertising index; wherein each advertising index corresponds to one piece of advertising contact data; the target advertising contact data contains a device identifier; A configuration module, used to configure a target advertisement attribution model based on the attribution strategy; The attribution module is used to input the target advertising contact data into the target advertising attribution model to obtain an advertising attribution result; the advertising attribution result includes at least one piece of advertising contact data.

10. The device according to claim 9, characterized in that The configuration module includes: A configuration rule generating unit, configured to generate configuration rules based on the attribution strategy; the configuration rules include filtering rules and processing rules; An execution process building unit, used to build an attribution execution process based on the configuration rules; A model configuration unit is used to configure the attribution execution process into a target advertising attribution model.

Images