Application download path determination method and apparatus, server, and storage medium

Abstract

Embodiments of the present disclosure provide a method and device for determining an application download path, a server and a storage medium. The method comprises: determining a device group associated with a download path of a target application; determining a target time window according to at least one time window corresponding to the device group; obtaining attribution parameter data of each model in a plurality of download path attribution models in the target time window, the attribution parameter data being used to represent a matching degree between a download path determined by the model and an actual download path of the target application; and determining a target model from the plurality of download path attribution models according to the attribution parameter data of each model in the target time window. Therefore, embodiments of the present disclosure can obtain a target model that more accurately reflects a real download path result.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a method, device, server, and storage medium for determining an application download path. Background Art

[0002] Users may download applications (APPs) through various paths. Currently, it is necessary to determine which download paths the users have taken. Using fingerprint links to determine the download paths of applications is a commonly used method in the industry.

[0003] However, the accuracy of the download path results obtained by the current method of using fingerprint links to determine the download path of an application is low. Summary of the Invention

[0004] The present disclosure provides a method, device, server, and storage medium for determining an application download path to at least address the low accuracy of download path results obtained using existing technologies. The technical solutions of the present disclosure are as follows:

[0005] According to a first aspect of an embodiment of the present disclosure, a method for determining an application download path is provided, the method comprising:

[0006] Determine the device group associated with the download path of the target application;

[0007] determining a target time window according to at least one time window corresponding to the device group;

[0008] Obtaining attribution parameter data for each of a plurality of download path attribution models within the target time window, the attribution parameter data being used to characterize a degree of match between a download path determined by the model and an actual download path of the target application;

[0009] A target model is determined from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

[0010] In one possible implementation, determining the device group associated with the download path of the target application includes:

[0011] Obtaining historical download path characteristics and newly added download path characteristics of the target application;

[0012] A first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path are determined, and the device groups associated with the download path are the first device group and the second device group.

[0013] In one possible implementation, determining the target time window based on at least one time window corresponding to the device group includes:

[0014] determining a first time window corresponding to the first device group and a second time window corresponding to the second device group;

[0015] A target time window is determined according to an overlapping time window of the first time window and the second time window.

[0016] In one possible implementation, determining the first time window corresponding to the first device group includes:

[0017] Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold;

[0018] A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

[0019] In one possible implementation, before determining the second time window corresponding to the second device group, the method further includes:

[0020] Obtain the type of the newly added download path feature;

[0021] Then, determining the second time window corresponding to the second device group includes:

[0022] calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature;

[0023] The time window corresponding to the highest first coincidence degree is determined as the second time window.

[0024] In one possible implementation, obtaining attribution parameter data of each of the multiple download path attribution models in the target time window includes:

[0025] Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application;

[0026] Determine a second download path corresponding to the behavioral characteristic parameters of each model;

[0027] Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model;

[0028] The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

[0029] In one possible implementation, determining the second download path corresponding to the behavioral characteristic parameters of each model includes:

[0030] calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model;

[0031] Determine a second download path corresponding to the second degree of overlap of each model;

[0032] The behavioral characteristic parameters are the user's retention time and usage time on the target application.

[0033] In one possible implementation, determining a target model from the multiple download path attribution models based on the attribution parameter data of each model in the target time window includes:

[0034] Calculate the harmonic mean of the attribution parameter data of each model in the target time window;

[0035] The model corresponding to the harmonic mean within a preset range is determined as the target model.

[0036] In one possible implementation, the download path features include at least one of the following: an identification number (ID) of a device used in the download path, a program package used by a user to download and install an application, an Internet Protocol User Agent (IPUA) obtained by a user by operating an off-site Internet Hypertext Markup Language (HTML5) page, and a password / invitation code match.

[0037] According to a second aspect of an embodiment of the present disclosure, a device for determining an application download path is provided, the device comprising:

[0038] A first determining module is configured to determine a device group associated with a download path of a target application;

[0039] a second determining module configured to determine a target time window according to at least one time window corresponding to the device group;

[0040] A first acquisition module is configured to obtain attribution parameter data of each model in a plurality of download path attribution models within the target time window, wherein the attribution parameter data is used to represent a matching degree between the download path determined by the model and the actual download path of the target application;

[0041] The third determination module is configured to determine a target model from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

[0042] In a possible implementation manner, the first determining module is specifically configured to:

[0043] Obtaining historical download path characteristics and newly added download path characteristics of the target application;

[0044] A first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path are determined, and the device groups associated with the download path are the first device group and the second device group.

[0045] In one possible implementation, when the second determination module determines the target time window according to the at least one time window corresponding to the device group, it is specifically configured to:

[0046] determining a first time window corresponding to the first device group and a second time window corresponding to the second device group;

[0047] A target time window is determined according to an overlapping time window of the first time window and the second time window.

[0048] In a possible implementation manner, when determining the first time window corresponding to the first device group, the second determination module is specifically configured to:

[0049] Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold;

[0050] A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

[0051] In one possible implementation, before determining the second time window corresponding to the second device group, the second determining module is further specifically configured to:

[0052] Obtain the type of the newly added download path feature;

[0053] Then, when determining the second time window corresponding to the second device group, the second determining module is specifically configured to:

[0054] calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature;

[0055] The time window corresponding to the highest first coincidence degree is determined as the second time window.

[0056] In a possible implementation manner, the first acquisition module is specifically configured to:

[0057] Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application;

[0058] Determine a second download path corresponding to the behavioral characteristic parameters of each model;

[0059] Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model;

[0060] The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

[0061] In one possible implementation, when determining the second download path corresponding to the behavioral characteristic parameter of each model, the first acquisition module is specifically configured to:

[0062] calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model;

[0063] Determine a second download path corresponding to the second degree of overlap of each model;

[0064] The behavioral characteristic parameters are the user's retention time and usage time on the target application.

[0065] In a possible implementation manner, the third determining module is specifically configured to:

[0066] Calculate the harmonic mean of the attribution parameter data of each model in the target time window;

[0067] The model corresponding to the harmonic mean within a preset range is determined as the target model.

[0068] In one possible implementation, the download path features include at least one of the following: an identification number (ID) of a device used in the download path, a program package used by a user to download and install an application, an Internet Protocol User Agent (IPUA) obtained by a user by operating an off-site Internet Hypertext Markup Language (HTML5) page, and a password / invitation code match.

[0069] According to a third aspect of an embodiment of the present disclosure, a server is provided, the server including:

[0070] processor;

[0071] a memory for storing instructions executable by the processor;

[0072] The processor is configured to execute the instructions to implement the application download path determination method provided in the present disclosure.

[0073] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of a server, the server is enabled to execute the method for determining an application download path described in the first aspect above.

[0074] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the application download path determination method provided by the present disclosure.

[0075] The technical solutions provided by the embodiments of the present disclosure bring at least the following beneficial effects:

[0076] The technical solution provided by the embodiments of the present disclosure determines the device group associated with the download path of the target application, and determines the target time window based on at least one time window corresponding to the device group, that is, obtaining a target time window with a higher recall rate through the at least one time window corresponding to the device group; obtaining attribution parameter data of each model in the target time window for each model of multiple download path attribution models, the attribution parameter data is used to characterize the degree of match between the download path determined by the model and the actual download path of the target application, that is, obtaining the degree of match between the download path determined by each model in the target time window and the actual download path, thereby determining the accuracy of the application download path obtained by each model in the target time window based on the degree of match; then, determining the target model from the multiple download path attribution models based on the attribution parameter data of each model in the target time window, that is, selecting the optimal model as the final download path attribution model based on the accuracy of the application download path obtained by each model in the target time window. Using this model can more accurately reflect the actual download path results.

[0077] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute an improper limitation of the present disclosure.

[0079] Figure 1 is a flowchart of a method for determining an application download path according to an exemplary embodiment;

[0080] Figure 2is a schematic diagram of the relationship between the determined download path and the actual download path according to an exemplary embodiment;

[0081] Figure 3 is a flowchart of another application download path determination method according to an exemplary embodiment;

[0082] Figure 4 is a block diagram of an application download path determination apparatus according to an exemplary embodiment;

[0083] Figure 5 is a block diagram of a server according to an exemplary embodiment. DETAILED DESCRIPTION

[0084] In order for those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings.

[0085] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented in an order other than that illustrated or described herein. The implementation described in the following exemplary embodiments does not represent all implementations consistent with the present disclosure. Rather, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims. In addition, the user information (including but not limited to user device information, user personal information, etc.) and related data involved in the present disclosure are information authorized by the user or authorized by each party.

[0086] Figure 1 is a flowchart of an application download path determination method according to an exemplary embodiment, as shown in Figure 1 The application download path determination method is adapted to electronic devices such as terminals, servers, etc., Figure 1 The embodiments shown in

[0087] In step S11, the device group associated with the download path of the target application is determined.

[0088] Specifically, the download path of the target application is obtained, and the device group associated with the download path of the target application is determined, wherein the device group associated with the download path of the target application refers to at least one device that installs the target application through the download path.

[0089] Among them, obtaining the download path of the target application is equivalent to attributing the download source of the target application, that is, determining through various technical means which channels the download of the target application installed on each device in the device group comes from, such as: the download link of the target application carried in other applications, the target application keyword search in the software store, etc.

[0090] In step S12, a target time window is determined according to at least one time window corresponding to the device group.

[0091] Specifically, a device group may correspond to at least one time range, i.e., at least one time window. Based on this at least one time window, a target time window can be derived. This target time window can be one of the one or more time windows, a combination of the aforementioned time windows, or a repeated combination of the aforementioned time windows, without specific limitations. The target time window is a time range with a high recall rate.

[0092] In step S13, attribution parameter data of each model in the plurality of download path attribution models in the target time window is obtained, and the attribution parameter data is used to characterize the matching degree between the download path determined by the model and the actual download path of the target application.

[0093] Specifically, by determining a target time window with a high recall rate, we can obtain the attribution parameter data for each download path attribution model within that target time window. This means we can determine the degree of match between the download path determined by each download path attribution model within that target time window and the actual download path of the target application. In other words, the accuracy of the download path determined by each model within that target time window with a high recall rate and the actual download path can be used to assess the degree of match. Specifically, if the download path determined by each model within the target time window has a high accuracy rate compared to the actual download path, then their degree of match is high; if the download path determined by each model has a low accuracy rate compared to the actual download path, then their degree of match is low.

[0094] The degree of match between the download path determined by each model within the target time window and the actual download path can be specifically expressed as the accuracy of the download path determined by each download path attribution model within the target time window with a high recall rate compared to the actual download path. A high degree of match indicates that the model accurately determined the download path within the target time window; a low degree of match indicates that the model accurately determined the download path within the target time window.

[0095] The download path attribution model can be a linear attribution model, a time-decay attribution model, or other models. The linear attribution model assigns equal weight to all channels corresponding to the download path. For example, if the download path includes channels a and b, then channel a has a weight of 50% and channel b also has a weight of 50%. The time-decay attribution model assigns a higher weight to the channels corresponding to the download path. For example, if the download path includes channels c and d, and channel a has a shorter time to conversion than channel b, then channel a has a weight of 70% and channel b has a weight of 30%. It should be understood that the channel corresponding to the download path refers to the source channel of the download path. For example, if application A on a device is downloaded using the download link of the target application, then the channel corresponding to the download path of application A on the device is the download link of the target application.

[0096] For example: Figure 2 As shown, if the matching degree between the download path determined by each model and the actual download path is specifically expressed by accuracy and recall, then the above accuracy = D / (D+E), that is, the overlap D between the data results calculated under the model (i.e., the download path E determined by the model) and the actual user source channel (i.e., the actual download path F) is divided by the total amount of model data.

[0097] Recall rate = D / (D+F), that is, the overlap D between the data results calculated under the model (i.e., the download path E determined by the model) and the actual user source channel (i.e., the actual download path F) divided by the total number of actual user source channels.

[0098] In step S14, a target model is determined from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

[0099] Specifically, based on the attribution parameter data of each download path attribution model in the target time window, it can be determined which download path attribution model can improve the attribution precision (i.e., accuracy and recall), that is, the download path attribution model is determined as the final target model.

[0100] For example, the download path of model A in the target time window is channel A, and the download path of model B in the target time window is channel B. If the accuracy of channel A is higher than that of channel B, and the recall rate of channel A is higher than that of channel B, then channel A can be determined to be the final download path. Therefore, model A is the target model, and using this target model for attribution can more accurately reflect the actual download path results.

[0101] As can be seen from the above steps S11 to S14, an embodiment of the present disclosure determines a device group associated with the download path of a target application, and determines a target time window based on at least one time window corresponding to the device group, that is, obtaining a target time window with a higher recall rate based on the at least one time window corresponding to the device group; obtaining attribution parameter data for each model in the target time window for each model of multiple download path attribution models, the attribution parameter data is used to characterize the degree of match between the download path determined by the model and the actual download path of the target application, that is, obtaining the degree of match between the download path determined by each model in the target time window and the actual download path, thereby determining the accuracy of the application download path obtained by each model in the target time window based on the degree of match, and thereby evaluating whether a download path attribution model is qualified; then, determining a target model from the multiple download path attribution models based on the attribution parameter data of each model in the target time window, that is, selecting the optimal model as the final download path attribution model based on the accuracy of the application download path obtained by each model in the target time window, which can more accurately reflect the actual download path result, that is, using a qualified model to obtain the download path can more accurately reflect the actual download path.

[0102] Figure 3 FIG. 1 is a specific flow chart of a method for determining an application download path according to an exemplary embodiment. Figure 3 As shown, the application download path determination method is applicable to electronic devices, such as terminals, servers, etc. Figure 3 The illustrated embodiment includes at least the following steps S21-S25.

[0103] Step S21: Acquire historical download path features and newly added download path features of the target application.

[0104] Specifically, the historical download path feature can be a historical attribution fingerprint, and the newly added download path feature can be a new attribution fingerprint. In the above steps, the historical download path feature of the target application is obtained, that is, the historical attribution fingerprint of the target application is obtained, and the newly added download path feature of the target application is obtained, that is, the newly added new attribution fingerprint of the target application is obtained.

[0105] Download path features (including historical download path features and newly added download path features) include but are not limited to at least one of the following: the International Mobile Equipment Identity (IMEI), Open Anonymous Device Identifier (OAID), Identifier for Advertising (IDFA), and other unique device identifiers used in the download path; the application package (APP) downloaded and installed by the user, such as the Android application package (APK); the Internet Protocol User-Agent (IP UA) fingerprint obtained by the user by operating an off-site Internet Hypertext Markup Language (Hyml5) page; and password / invitation code matching. The above download path features can all be used as fingerprints for attribution.

[0106] Step S22: Determine a first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path, where the device groups associated with the download path are the first device group and the second device group.

[0107] Specifically, after obtaining the historical download path characteristics of the target application, a first device group corresponding to the historical download path characteristics (i.e., the device group that has historically used the target application) is determined. This first device group is one of the device groups associated with the download path. Furthermore, after obtaining the newly added download path characteristics, a second device group corresponding to the newly added download path characteristics (i.e., the newly added device group that has used the target application) is determined. This second device group is one of the device groups associated with the download path.

[0108] Step S23: determining a target time window according to at least one time window corresponding to the device group.

[0109] Specifically, the device group may correspond to at least one time range, that is, at least one time window. A target time window can be obtained based on the above at least one time window. The target time window can be one of the above one or more time windows, or a time window combined with the above multiple time windows, or a repeated time window of the above multiple time windows. No specific limitation is made here.

[0110] The target time window is a time range with a higher recall rate. If the recall rate of one of the one or more time windows is higher than that of the other time windows, and the combined time window and the repeated time window, the time window with the highest recall rate is selected as the target time window; if the recall rate of the combined time window of the multiple time windows is the highest, the combined time window is selected as the target time window; if the recall rate of the repeated time window of the multiple time windows is the highest, the repeated time window is selected as the target time window.

[0111] Step S24: Obtain the attribution parameter data of each model in the multiple download path attribution models in the target time window, the attribution parameter data being used to represent the matching degree of the download path determined by the model and the actual download path of the target application.

[0112] Specifically, since the target time window with a higher recall rate is determined, the attribution parameter data of each download path attribution model in the target time window can be obtained, that is, the matching degree of the download path determined by each download path attribution model in the target time window and the actual download path of the target application can be obtained, that is, the accuracy of the download path determined by each model in the target time window with a higher recall rate and the actual download path can be used to judge the matching degree. That is, if the accuracy of the download path determined by each model in the target time window and the actual download path is higher, it means that the matching degree is higher; if the accuracy of the download path determined by each model and the actual download path is lower, it means that the matching degree is lower.

[0113] The matching degree of the download path determined by each model in the target time window and the actual download path can be specifically represented as: the accuracy of the download path determined by each download path attribution model in the target time window with a higher recall rate compared with the actual download path. If the matching degree is high, it means that the accuracy of the download path determined by the model in the target time window is high; if the matching degree is low, it means that the accuracy of the download path determined by the model in the target time window is low.

[0114] The download path attribution model can be a linear attribution model, a time-decay attribution model, or other models. The linear attribution model assigns equal weight to all channels corresponding to the download path. For example, if the download path includes channels a and b, then channel a has a weight of 50% and channel b also has a weight of 50%. The time-decay attribution model assigns a higher weight to the channels corresponding to the download path. For example, if the download path includes channels c and d, and channel a has a shorter time to conversion than channel b, then channel a has a weight of 70% and channel b has a weight of 30%. It should be understood that the channel corresponding to the download path refers to the source channel of the download path. For example, if application A on a device is downloaded using the download link of the target application, then the channel corresponding to the download path of application A on the device is the download link of the target application.

[0115] Step S25: determining a target model from the plurality of download path attribution models according to the attribution parameter data of each model in the target time window.

[0116] Specifically, based on the attribution parameter data of each download path attribution model in the target time window, it can be determined which download path attribution model can improve the attribution precision (i.e., accuracy and recall), that is, the download path attribution model is determined as the final target model.

[0117] For example, the download path of model A in the target time window is channel A, and the download path of model B in the target time window is channel B. If the accuracy of channel A is higher than that of channel B, and the recall rate of channel A is higher than that of channel B, then channel A can be determined to be the final download path. Therefore, model A is the target model, and using this target model for attribution can more accurately reflect the actual download path results.

[0118] As can be seen from the above steps S21 to S25, in an embodiment of the present disclosure, by obtaining the historical download path characteristics and the newly added download path characteristics of the target application, and determining the first device group corresponding to the historical download path characteristics and the second device group corresponding to the newly added download path, a target time window is determined according to at least one time window corresponding to the device groups (i.e., the first device group and the second device group), that is, a target time window with a higher recall rate is obtained through at least one time window corresponding to the device group; attribution parameter data of each model in multiple download path attribution models under the target time window is obtained, and the attribution parameter data is used to The matching degree between the download path determined by the characterization model and the actual download path of the target application is obtained, that is, the matching degree between the download path determined by each model in the target time window and the actual download path is obtained, thereby the accuracy of the application download path obtained by each model in the target time window can be known through the matching degree; then, according to the attribution parameter data of each model in the target time window, the target model is determined from the multiple download path attribution models, that is, the accuracy of the application download path obtained by each model in the target time window is used to select the optimal model as the final download path attribution model. The use of this model can more accurately reflect the actual download path results.

[0119] In a possible implementation, step S12 determines the target time window according to at least one time window corresponding to the device group, specifically including:

[0120] determining a first time window corresponding to the first device group and a second time window corresponding to the second device group;

[0121] A target time window is determined according to an overlapping time window of the first time window and the second time window.

[0122] Specifically, after determining the first device group corresponding to the historical download path characteristics, a time range corresponding to the first device group is determined, i.e., a first time window in which the first device group is relatively concentrated. It should be understood that after determining the second device group corresponding to the newly added download path, a time range corresponding to the second device group is determined, i.e., a second time window in which the second device group is relatively concentrated.

[0123] Based on the time ranges of the first time window and the second time window, it is determined whether the two time ranges overlap at least partially. If at least partially overlap, the overlapping time range between the first time window and the second time window is used as the target time window. If there is no overlap between the first time window and the second time window, the first time window can be used as the target time window, the second time window can be used as the target time window, or both the first time window and the second time window can be used as the target time window, without specific limitation.

[0124] It should be noted that the steps of determining the first time window and the steps of determining the second time window are not limited to a certain order. The second time window can be determined first and then the first time window, or the first time window can be determined first and then the second time window. It can be set as needed and is not specifically limited here.

[0125] In a possible implementation, the step of determining the first time window corresponding to the first device group may specifically include:

[0126] Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold;

[0127] A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

[0128] Specifically, we analyze the impact of time windows of different granularities on the increase in attribution magnitude, that is, obtain the time inflection points of the high-speed conversion time window and the climbing conversion time window, and then determine the time range with a higher recall rate, that is, the first time window, based on the time inflection point.

[0129] It should be noted that the method for determining whether the recall rate is high can be determined by determining whether the growth rate of the first device group exceeds a first threshold. If the growth rate of the first device group exceeds the first threshold, the time range is determined to be high. If the growth rate of the first device group is below the first threshold, the time range is determined to be low. The high-speed conversion time window is the time range with a high recall rate, and the ramp-up conversion time window is the time range with a low recall rate.

[0130] In a possible implementation, before determining the second time window corresponding to the second device group, the method may further include:

[0131] Obtain the type of the newly added download path feature;

[0132] Then, the step of determining the second time window corresponding to the second device group may specifically include:

[0133] calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature;

[0134] The time window corresponding to the highest first coincidence degree is determined as the second time window.

[0135] Specifically, the type of the newly added download path feature is obtained. After obtaining the type of the newly added download path feature, the first overlap between the second device group and the third device group corresponding to the same high-precision download path feature (i.e., historical download path features of the same type as the newly added download path feature) is calculated under different time windows. Based on the first overlap calculated under different time windows, a reasonable time window for the newly added download path feature can be determined, that is, the time window with the highest first overlap is used as the second time window. The first overlap is the number or ratio of devices that overlap between the second device group and the third device group.

[0136] In a possible implementation, step S13 of obtaining attribution parameter data of each of the multiple download path attribution models in the target time window may specifically include:

[0137] Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application;

[0138] Determine a second download path corresponding to the behavioral characteristic parameters of each model;

[0139] Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model;

[0140] The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

[0141] Specifically, the first download path of each model in the target time window, i.e., the source channel, etc., is obtained; and the behavioral characteristic parameters corresponding to each model are obtained, and the second download path corresponding to the behavioral characteristic parameters corresponding to each model is determined, that is, the behavioral characteristic parameters corresponding to each model can obtain the corresponding download path through the model, i.e., the second download path.

[0142] A determination is made as to whether the first and second download paths of each model overlap. If so, the overlapping download path is determined as the download path determined by the model. The degree of match between the download path determined by each model and the actual download path is then used as the attribution parameter data for the model within the target time window. If there are no overlapping download paths, the first download path can be set as the download path determined by the model, or the second download path can be set as the download path determined by the model, etc., as needed, without specific limitations here.

[0143] The behavior characteristic parameters may specifically be parameters related to the user's usage habits of the target application.

[0144] It should be noted that the steps of obtaining the first download path and obtaining the second download path are not limited to the order of sequence. It is also possible to first obtain the behavioral characteristic parameters and further determine the second download path, and then obtain the first download path; or it is possible to first obtain the first download path and the behavioral characteristic parameters, and then determine the second download path based on the behavioral characteristic parameters. No specific limitation is made here.

[0145] In a possible implementation, the step of determining the second download path corresponding to the behavioral characteristic parameters of each model may specifically include:

[0146] calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model;

[0147] Determine a second download path corresponding to the second degree of overlap of each model;

[0148] The behavioral characteristic parameters are the user's retention time and usage time on the target application.

[0149] Specifically, first, the historical download path characteristics / channels with higher conversion efficiency are used as the control group, and the low-efficiency download path characteristics or newly added download path characteristics are used as the experimental group. The overlap between the first parameter of the behavioral characteristics of the first device group for the target application and the second parameter of the behavioral characteristics of the second device group for the target application of each model is calculated. In this way, the second download path corresponding to the second overlap of each model can be further determined, that is, the corresponding download path can be obtained through the second overlap of each model.

[0150] In a possible implementation, step S14 determines the target model from the multiple download path attribution models based on the attribution parameter data of each model in the target time window, which may specifically include:

[0151] Calculate the harmonic mean of the attribution parameter data of each model in the target time window;

[0152] The model corresponding to the harmonic mean within a preset range is determined as the target model.

[0153] Specifically, the harmonic mean of the attribution parameter data for each model within the target time window is calculated. Whether the harmonic mean falls within a preset range determines whether it can be used as the final target model. If the harmonic mean falls within the preset range, the model corresponding to the harmonic mean can be used as the final target model; if the harmonic mean does not fall within the preset range, the model corresponding to the harmonic mean cannot be used as the final target model. Harmonic mean = precision * recall * 2 / (precision + recall).

[0154] It should be noted that a preset range of the harmonic mean may be set as needed. If the harmonic mean is within the preset range, the model corresponding to the harmonic mean is determined to be the target model.

[0155] To sum up, the present disclosure mainly obtains the historical download path characteristics and the newly added download path characteristics of the target application, and determines the first device group corresponding to the historical download path characteristics and the second device group corresponding to the newly added download path. According to at least one time window corresponding to the device group, a target time window with a higher recall rate is determined. By obtaining the matching degree between the download path determined by each model in the target time window with a higher recall rate and the actual download path, the optimal model is selected as the final download path attribution model. The use of this model can more accurately reflect the actual download path results.

[0156] Figure 4 FIG. 1 is a block diagram of a device for determining an application download path according to an exemplary embodiment. Figure 4 The application download path determination device 40 includes a first determination module 41 , a second determination module 42 , a first acquisition module 43 and a third determination module 44 .

[0157] A first determining module 41 is configured to determine a device group associated with a download path of a target application;

[0158] A second determining module 42 is configured to determine a target time window according to at least one time window corresponding to the device group;

[0159] A first acquisition module 43 is configured to obtain attribution parameter data of each model in the plurality of download path attribution models within the target time window, wherein the attribution parameter data is used to represent a degree of match between the download path determined by the model and the actual download path of the target application;

[0160] The third determining module 44 is configured to determine a target model from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

[0161] In a possible implementation, the first determining module 41 is specifically configured to:

[0162] Obtaining historical download path characteristics and newly added download path characteristics of the target application;

[0163] A first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path are determined, and the device groups associated with the download path are the first device group and the second device group.

[0164] In a possible implementation, when determining the target time window according to the at least one time window corresponding to the device group, the second determination module 42 is specifically configured to:

[0165] determining a first time window corresponding to the first device group and a second time window corresponding to the second device group;

[0166] A target time window is determined according to an overlapping time window of the first time window and the second time window.

[0167] In a possible implementation, when determining the first time window corresponding to the first device group, the second determining module 42 is specifically configured to:

[0168] Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold;

[0169] A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

[0170] In a possible implementation manner, before determining the second time window corresponding to the second device group, the second determining module 42 is further specifically configured to:

[0171] Obtain the type of the newly added download path feature;

[0172] Then, when determining the second time window corresponding to the second device group, the second determining module 42 is specifically configured to:

[0173] calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature;

[0174] The time window corresponding to the highest first coincidence degree is determined as the second time window.

[0175] In a possible implementation, the first acquisition module 43 is specifically configured to:

[0176] Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application;

[0177] Determine a second download path corresponding to the behavioral characteristic parameters of each model;

[0178] Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model;

[0179] The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

[0180] In a possible implementation, when determining the second download path corresponding to the behavioral characteristic parameter of each model, the first acquisition module 43 is specifically configured as follows:

[0181] calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model;

[0182] Determine a second download path corresponding to the second degree of overlap of each model;

[0183] The behavioral characteristic parameters are the user's retention time and usage time on the target application.

[0184] In a possible implementation, the third determining module 44 is specifically configured to:

[0185] Calculate the harmonic mean of the attribution parameter data of each model in the target time window;

[0186] The model corresponding to the harmonic mean within a preset range is determined as the target model.

[0187] In a possible implementation, the download path features include at least one of the following: an identification number (ID) of a device used in the download path, a program package used by a user to download and install an application, an Internet Protocol User Agent (IPUA) obtained by a user by operating an off-site Internet Hypertext Markup Language (HTML5) page, and a password / invitation code match.

[0188] To sum up, the present disclosure mainly obtains the historical download path characteristics and the newly added download path characteristics of the target application, and determines the first device group corresponding to the historical download path characteristics and the second device group corresponding to the newly added download path. According to at least one time window corresponding to the device group, a target time window with a higher recall rate is determined. By obtaining the matching degree between the download path determined by each model in the target time window with a higher recall rate and the actual download path, the optimal model is selected as the final download path attribution model. The use of this model can more accurately reflect the actual download path results.

[0189] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0190] Figure 5 FIG. 1 is a block diagram of a server according to an exemplary embodiment. Figure 5 , the server includes:

[0191] Processor 510;

[0192] a memory 520 for storing instructions executable by the processor;

[0193] The processor is configured to execute the instructions to implement the above-mentioned method for determining the application download path applied to the server.

[0194] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory including instructions. The instructions can be executed by the processor 510 of the server to perform the above method. Alternatively, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0195] In an exemplary embodiment, a computer program product is further provided, including a computer program / instruction, which implements the above-mentioned method for determining an application download path when executed by a processor.

[0196] In addition, the application download path determination scheme provided herein is not inherently related to any specific computer, virtual system, or other device. Various general-purpose systems can also be used together with the teachings based on this. Based on the above description, it is obvious that the structure required for constructing a system with the scheme of the present invention is suitable. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to implement the content of the present invention described herein, and the above description of specific languages ​​is intended to disclose the best mode of implementation of the present invention.

[0197] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0198] Similarly, it should be understood that in order to streamline the present disclosure and aid understanding of one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the claims, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Accordingly, the claims that follow the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the invention.

[0199] Those skilled in the art will appreciate that the modules in the devices of the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device so disclosed may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0200] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims, any of the claimed embodiments may be used in any combination.

[0201] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It will be appreciated by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components in the information extraction scheme according to the disclosed embodiments. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and a computer program product) for executing a part or all of the methods described herein. Such a program implementing the present invention can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0202] It should be noted that the above embodiments illustrate rather than limit the invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising several different elements and by means of appropriately programmed computers. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0203] In summary, those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0204] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for determining an application download path, characterized in that: The method comprises: Determining a device group associated with a download path of a target application; the download path of the target application indicates a download source of the target application; the device group refers to at least one device that installs the target application through the download path; Determining a target time window based on at least one time window corresponding to the device group; the time window refers to a time range within which the device group installs the target application through the download path; Obtaining attribution parameter data for each of a plurality of download path attribution models within the target time window, the attribution parameter data being used to characterize a degree of match between a download path determined by the model and an actual download path of the target application; A target model is determined from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

2. The method for determining an application download path according to claim 1, wherein: Determining the device group associated with the download path of the target application includes: Obtaining historical download path characteristics and newly added download path characteristics of the target application; A first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path are determined, and the device groups associated with the download path are the first device group and the second device group.

3. The method for determining an application download path according to claim 2, wherein: The determining the target time window according to the at least one time window corresponding to the device group includes: determining a first time window corresponding to the first device group and a second time window corresponding to the second device group; A target time window is determined according to an overlapping time window of the first time window and the second time window.

4. The method for determining an application download path according to claim 3, wherein: The determining a first time window corresponding to the first device group includes: Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold; A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

5. The method for determining an application download path according to claim 3, wherein: Before determining the second time window corresponding to the second device group, the method further includes: Obtain the type of the newly added download path feature; Then, determining the second time window corresponding to the second device group includes: calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature; The time window corresponding to the highest first coincidence degree is determined as the second time window.

6. The method for determining an application download path according to claim 3, wherein: The obtaining of attribution parameter data of each model in the plurality of download path attribution models in the target time window includes: Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application; Determine a second download path corresponding to the behavioral characteristic parameters of each model; Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model; The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

7. The method for determining an application download path according to claim 6, wherein: The determining of the second download path corresponding to the behavioral characteristic parameters of each model includes: calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model; Determine a second download path corresponding to the second degree of overlap of each model; The behavioral characteristic parameters are the user's retention time and usage time on the target application.

8. The method for determining an application download path according to claim 1, wherein: Determining a target model from the plurality of download path attribution models according to the attribution parameter data of each model in the target time window includes: Calculate the harmonic mean of the attribution parameter data of each model in the target time window; The model corresponding to the harmonic mean within a preset range is determined as the target model.

9. The method for determining an application download path according to any one of claims 1 to 8, wherein: The characteristics of the download path include at least one of the following: the identity identification number ID of the device used in the download path, the program package of the user downloaded and installed the application, the Internet Protocol User Agent IPUA obtained by the user by operating the Internet Hypertext Markup Language HTML5 page outside the site, and the password / invitation code matching.

10. A device for determining an application download path, characterized in that: The device comprises: A first determining module is configured to determine a device group associated with a download path of a target application; the download path of the target application represents a download source of the target application; the device group refers to at least one device that installs the target application through the download path; A second determining module is configured to determine a target time window based on at least one time window corresponding to the device group; the time window refers to a time range within which the device group installs the target application through the download path; A first acquisition module is configured to obtain attribution parameter data of each model in a plurality of download path attribution models within the target time window, wherein the attribution parameter data is used to represent a matching degree between the download path determined by the model and the actual download path of the target application; The third determination module is configured to determine a target model from the multiple download path attribution models according to the attribution parameter data of each model in the target time window.

11. The device for determining an application download path according to claim 10, wherein: The first determining module is specifically configured to: Obtaining historical download path characteristics and newly added download path characteristics of the target application; A first device group corresponding to the historical download path characteristics and a second device group corresponding to the newly added download path are determined, and the device groups associated with the download path are the first device group and the second device group.

12. The device for determining an application download path according to claim 11, wherein: When determining the target time window according to the at least one time window corresponding to the device group, the second determining module is specifically configured to: determining a first time window corresponding to the first device group and a second time window corresponding to the second device group; A target time window is determined according to an overlapping time window of the first time window and the second time window.

13. The device for determining an application download path according to claim 12, wherein: When determining the first time window corresponding to the first device group, the second determining module is specifically configured to: Obtaining a time inflection point of the growth rate of the first device group; wherein the time inflection point is a time point when the growth rate reaches a first threshold; A time window in which the growth rate of the first device group exceeds the first threshold is determined as a first time window.

14. The device for determining an application download path according to claim 12, wherein: Before determining the second time window corresponding to the second device group, the second determining module is further specifically configured to: Obtain the type of the newly added download path feature; Then, when determining the second time window corresponding to the second device group, the second determining module is specifically configured to: calculating a first degree of overlap between the second device group and a third device group in different time windows, the third device group corresponding to historical download path features of the same type as the newly added download path feature; The time window corresponding to the highest first coincidence degree is determined as the second time window.

15. The device for determining an application download path according to claim 12, wherein: The first acquisition module is specifically configured to: Obtaining a first download path of each model in the target time window and behavior characteristic parameters corresponding to each model from a plurality of download path attribution models, the behavior characteristic parameters including: a first parameter of a behavior characteristic of the first device group with respect to the target application and a second parameter of a behavior characteristic of the second device group with respect to the target application; Determine a second download path corresponding to the behavioral characteristic parameters of each model; Obtaining a download path determined by the model according to the overlapping download path of the first download path and the second download path of each model; The matching degree between the download path determined by each model and the actual download path is determined as the attribution parameter data of the model in the target time window.

16. The device for determining an application download path according to claim 15, wherein: When determining the second download path corresponding to the behavioral characteristic parameter of each model, the first acquisition module is specifically configured to: calculating a second degree of coincidence between the first parameter and the second parameter in the behavioral characteristic parameters of each model; Determine a second download path corresponding to the second degree of overlap of each model; The behavioral characteristic parameters are the user's retention time and usage time on the target application.

17. The device for determining an application download path according to claim 10, wherein: The third determining module is specifically configured to: Calculate the harmonic mean of the attribution parameter data of each model in the target time window; The model corresponding to the harmonic mean within a preset range is determined as the target model.

18. The device for determining an application download path according to any one of claims 10 to 17, wherein: The characteristics of the download path include at least one of the following: the identity identification number ID of the device used in the download path, the program package of the user downloaded and installed the application, the Internet Protocol User Agent IPUA obtained by the user by operating the Internet Hypertext Markup Language HTML5 page outside the site, and the password / invitation code matching.

19. A server, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for determining an application download path according to any one of claims 1 to 9.

20. A computer-readable storage medium, characterized in that When the instructions in the computer-readable storage medium are executed by a processor of a server, the server is enabled to execute the application download path determination method according to any one of claims 1 to 9.

21. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method for determining an application download path according to any one of claims 1 to 9 is implemented.

Images