Equipment detection method and device, electronic equipment and storage medium

The device detection method enhances the identification of cheating devices by analyzing activation behavior data through feature calculation and time-based corrections, improving the accuracy of identifying and preventing fraudulent activities in application promotion plans.

CN119089265AActive Publication Date: 2024-12-06BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411132901.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-12-06
Estimated Expiration
2044-08-16

AI Technical Summary

Technical Problem

The growth of APP activation behavior is affected by cheating devices that perform machine brushing, leading to inaccurate promotion plan adjustments and disrupting the mobile ecology of applications.

Method used

A device detection method that involves acquiring activation behavior data, performing feature calculation on data slices, correcting features based on time distances, and using a cheating discrimination model to determine if a device is an application activation cheating device.

Benefits of technology

Improves the accuracy of identifying cheating devices by accurately determining the features of activation behavior data, enabling precise identification and prevention of fraudulent activities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089265A_ABST
    Figure CN119089265A_ABST
Patent Text Reader

Abstract

The invention provides an equipment detection method and device, electronic equipment and a storage medium, relates to the technical field of artificial intelligence, in particular to the fields of flow data processing, network ecological security, deep learning and the like, and can be applied to scenes such as search, social contact, shopping, travel, question and answer, local life, video and the like. The specific implementation scheme is as follows: acquiring activation behavior data of an application program in first equipment; performing feature calculation on each behavior data fragment in the activated behavior data to obtain a first feature of each behavior data fragment; based on the time distance between the behavior data fragments, performing feature correction on the first feature of each behavior data fragment to obtain a second feature of each behavior data fragment; and based on the second feature of each behavior data fragment, determining whether the first device is an application program activation cheating device.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, especially to the fields of traffic data processing, network ecological security and deep learning, and can be applied to various business scenarios based on mobile terminal devices such as search, social networking, shopping, travel, question and answer, local life, video, etc. The present disclosure specifically relates to a device detection method, apparatus, electronic device and storage medium. Background Art

[0002] On the Internet, some companies or businesses will promote some new APPs (Applications), such as search, social, shopping, travel, Q&A, local life, video, etc. At the same time, the growth of APP activation behavior in various channels will be recorded, and subsequent promotion plans will be adjusted based on the growth of each channel. However, some black and gray industries use cheating devices to perform machine brushing on APP activation behavior, which will affect the adjustment of APP promotion plans and destroy the mobile ecology of the APP. Summary of the invention

[0003] The present disclosure provides a device detection method, apparatus, electronic device and storage medium.

[0004] According to one aspect of the present disclosure, there is provided a device detection method, comprising:

[0005] Acquire activation behavior data of an application in the first device;

[0006] Performing feature calculation on each behavior data slice in the activation behavior data to obtain a first feature of each behavior data slice;

[0007] Based on the time distance between each of the behavior data slices, feature correction is performed on the first feature of each of the behavior data slices to obtain the second feature of each of the behavior data slices;

[0008] Based on the second characteristics of each of the behavior data slices, determine whether the first device is an application activation cheating device.

[0009] According to another aspect of the present disclosure, there is provided a device detection apparatus, comprising:

[0010] A data acquisition module, used to acquire activation behavior data of an application in the first device;

[0011] A feature calculation module, used for performing feature calculation on each behavior data slice in the activation behavior data to obtain a first feature of each behavior data slice;

[0012] A feature correction module, used for correcting the first feature of each of the behavior data slices based on the time distance between each of the behavior data slices to obtain a second feature of each of the behavior data slices;

[0013] The cheating detection module is used to determine whether the first device is an application-activated cheating device based on the second feature of each of the behavior data slices.

[0014] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0015] at least one central processing unit; and

[0016] A memory in communication with the at least one central processing unit; wherein,

[0017] The memory stores instructions that can be executed by the at least one central processing unit, and the instructions are executed by the at least one central processing unit so that the at least one central processing unit can execute any parking method in the embodiments of the present disclosure.

[0018] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any device detection method in the embodiments of the present disclosure.

[0019] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which implements any device detection method according to the embodiments of the present disclosure when executed by a processor.

[0020] According to the technology disclosed in the present invention, the activation behavior data of the application in the device is feature-calculated to obtain the first feature, and then the feature of each segment is corrected by using the influence of the time distance between each behavior data segment in the activation behavior data on the feature of each segment to obtain the second feature of each segment. In this way, the feature describing the activation behavior data can be accurately obtained, and then the feature can be identified to accurately determine whether the device is an activated cheating device. Thus, the accuracy of identifying activated cheating devices is improved.

[0021] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0023] Figure 1is a flow chart of a device detection method according to an embodiment of the present disclosure;

[0024] Figure 2 is a flow chart of a feature processing process of an embodiment of the present disclosure;

[0025] Figure 3 is a flow chart of an anti-cheating processing method according to an embodiment of the present disclosure;

[0026] Figure 4 is a structural block diagram of a device detection apparatus according to an embodiment of the present disclosure;

[0027] Figure 5 is a structural block diagram of a device detection apparatus according to another embodiment of the present disclosure;

[0028] Figure 6 is a block diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0030] In order to facilitate understanding of the device detection method provided by the embodiment of the present disclosure, the related technologies of the embodiment of the present disclosure are explained below. The following related technologies can be arbitrarily combined with the technical solutions of the embodiment of the present disclosure as optional solutions, and they all belong to the protection scope of the embodiment of the present disclosure.

[0031] Figure 1 : is a flow chart of a device detection method of an embodiment of the present disclosure. The method can be applied to an electronic device. The electronic device is, for example, a terminal, a server, or other processing device, wherein the terminal can be a user equipment (UE) such as a desktop computer, a mobile device, a PDA (Personal Digital Assistant), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some implementations, the electronic device can implement the device detection method of an embodiment of the present disclosure by a central processing unit calling a computer-readable instruction stored in a memory.

[0032] like Figure 1 As shown, the method may include the following steps:

[0033] S110, obtaining activation behavior data of an application in a first device;

[0034] S120, performing feature calculation on each behavior data slice in the activated behavior data to obtain a first feature of each behavior data slice;

[0035] S130, based on the time distance between each behavior data slice, modify the first feature of each behavior data slice to obtain a second feature of each behavior data slice;

[0036] S140: Determine whether the first device is an application activation cheating device based on the second characteristics of each behavior data segment.

[0037] It is understandable that a data collection module is built into the application. When the first device downloads the application and activates the application, the data collection module uses the activation behavior data at this time and uploads it to the cloud, which can store the activation behavior data and perform the above steps S110 to S140.

[0038] Exemplarily, the activation behavior data may be activation information and device information etc. triggered by the user in the application within a short period of time after the first device downloads the application and opens the application.

[0039] Exemplarily, the activation behavior data may be time series data or streaming data.

[0040] Exemplarily, the activation behavior data may include a plurality of behavior data segments, and the segments are arranged or connected to each other in chronological order.

[0041] Exemplarily, the first feature may include the occurrence frequency of each dimensional object under each dimensional category in multiple dimensional categories in the behavior data shard.

[0042] Exemplarily, the second feature may include the occurrence frequency of each dimensional object under each dimensional category in multiple dimensional categories in the behavior data shard.

[0043] Exemplarily, performing feature calculation on a shard includes: performing feature calculation on the shard according to the feature calculation formula corresponding to each dimension category in multiple dimension categories, and obtaining the features of the shard under each dimension category. For example, there may be multiple dimension objects under each dimension category, and feature statistics are performed on the frequency of occurrence of all dimension objects under the category in the shard to obtain the features of the shard under the dimension category. For example, concentration features, proportion features, deduplication count features, etc.

[0044] It can be understood that the first feature and the second feature may include the above-mentioned concentration feature, proportion feature, deduplication count feature, etc.

[0045] It can be understood that the concentration feature may include a feature obtained by statistically analyzing the concentration of occurrence frequencies of multiple dimensional objects in a certain category, or may include these features in multiple categories.

[0046] It can be understood that the proportion feature can be a feature obtained by statistically analyzing the proportion of the target occurrence frequency that meets the preset conditions relative to the total occurrence frequency among the occurrence frequencies of multiple dimensional objects under a certain category. Alternatively, it can include these features under multiple categories.

[0047] It is understandable that the deduplication count feature may include the occurrence frequency of each dimensional object under a certain category, or may include these features under multiple categories.

[0048] Exemplarily, in the above step S130, based on the influence of the time distance between each behavior data slice on the first feature of each behavior data slice, the first feature of each behavior data slice is modified to obtain the second feature of each behavior data slice.

[0049] For example, by utilizing the influence of the interval between each second behavior data slice and the first behavior data slice on the first feature of the first behavior data slice, the first feature of the first behavior data slice is corrected to obtain the second feature of the first behavior data slice.

[0050] It can be understood that the first behavior data segment and the second behavior data segment are both segments of the above-mentioned activation behavior data.

[0051] It can be understood that the first feature after feature correction is the above-mentioned second feature.

[0052] For example, in the above step S140, the second feature of each behavior data segment is input into the cheating discrimination model to obtain a discrimination result output by the cheating discrimination model, which is used to characterize whether the first device is an application-activated cheating device. The cheating discrimination model is a trained model.

[0053] Exemplarily, the method may further include: adjusting the model parameters of the cheating discrimination model based on the discrimination result output by the cheating discrimination model and the labeling result of the first device to obtain a new cheating discrimination model, wherein the labeling result is used to indicate whether the first device is an application-activated cheating device.

[0054] Exemplarily, after the second feature is processed in breadth and depth, it is input into the cheating discrimination model to obtain the discrimination result output by the cheating discrimination model.

[0055] According to the above implementation, the activation behavior data of the application in the device is feature calculated to obtain the first feature, and then the feature of each segment is corrected by using the influence of the time distance between each behavior data segment in the activation behavior data on the feature of each segment to obtain the second feature of each segment, so that the feature describing the activation behavior data can be accurately obtained, and then the feature can be identified to accurately determine whether the device is an activation cheating device. Thus, the accuracy of identifying the activation cheating device is improved.

[0056] In one embodiment, based on the time distance between each behavior data slice, the first feature of each behavior data slice is corrected to obtain the second feature of each behavior data slice, including: based on the time distance between the first behavior data slice and each second behavior data slice, the first feature of each second behavior data slice is respectively attenuated to obtain the feature influence of each second behavior data slice on the first behavior data slice; based on the feature influence of each second behavior data slice on the first behavior data slice, the first feature of the first behavior data slice is corrected to obtain the second feature of the first behavior data slice.

[0057] It can be understood that the first behavior data segment is a segment in the above-mentioned activation behavior data. The second behavior data segment is also a segment in the above-mentioned activation behavior data that is not the first behavior data segment.

[0058] It can be understood that the first behavior data segment is determined from the activation behavior data, and then all or part of the remaining segments are used as the second behavior data segment. The time length of each segment is the same.

[0059] Exemplarily, the impact of the time distance between the first behavior data slice and each second behavior data slice on the first feature of the first behavior data may include: performing feature attenuation on the first feature of each second behavior data slice at the time distance between the first behavior data slice and each second behavior data slice, and obtaining the feature influence of each second behavior data slice on the first behavior data slice.

[0060] It can be understood that the closer the time distance between two slices is, the greater the feature influence between the two slices is, and the farther the time distance between the two slices is, the smaller the feature influence between the two slices is.

[0061] It can be understood that the first feature of the first behavior data slice is modified by using the sum or weighted sum of the feature influences of each second behavior data slice on the first behavior data slice to obtain the second feature of the first behavior data slice.

[0062] For example, the sum of the above feature influences is added to the first feature to obtain the second feature of the first behavior data slice.

[0063] For another example, the sum of the above feature influences is multiplied by a coefficient and then added to the first feature to obtain the second feature of the first behavior data slice.

[0064] According to the above implementation, the distance between each shard and the current shard can be used to determine the influence of each shard's feature on the current shard's feature, thereby using the influence of each shard's feature on the current shard's feature to correct the current shard's feature, and more accurate shard features can be obtained. Subsequently, when using this feature to identify whether a device is a cheating device, the recognition accuracy can be further improved.

[0065] In one embodiment, based on the time distance between the first behavior data slice and each second behavior data slice, feature attenuation is performed on the first feature of each second behavior data slice to obtain the feature influence of each second behavior data slice on the first behavior data slice, including: determining the degree of attenuation of the influence of the second behavior data slice on the first behavior data slice based on the time distance between the first behavior data slice and the second behavior data slice, and a preset attenuation coefficient; based on the influence attenuation degree, feature attenuating the first feature of the second behavior data slice to obtain the feature influence of the second behavior data slice on the first behavior data slice.

[0066] Exemplarily, each behavior data segment is sorted in chronological order, and the time distance between each segment can be determined based on the sorting position.

[0067] Exemplarily, the difference between the timestamps of the earliest data of each shard is used as the time distance between the shards.

[0068] Exemplarily, the difference between the timestamps of the last data of each shard is used as the time distance between the shards.

[0069] Exemplarily, by taking the time distance between the first behavior data segment and the second behavior data segment as the exponent and the attenuation coefficient as the base, an exponential function calculation is performed to obtain the attenuation degree of the influence of the second behavior data segment on the first behavior data segment.

[0070] Exemplarily, by taking the time distance between the first behavior data segment and the second behavior data segment as the independent variable and the attenuation coefficient as a fixed value, and multiplying them, the dependent variable, that is, the attenuation degree of the influence of the second behavior data segment on the first behavior data segment, can be obtained.

[0071] Exemplarily, the above-mentioned influence attenuation degree is multiplied by the first feature of the second behavior data slice to obtain the feature influence amount of the second behavior data slice on the first behavior data slice.

[0072] Exemplarily, another mapping value is determined based on the above-mentioned impact attenuation degree, and the mapping value is multiplied by the first feature of the second behavior data slice to obtain the feature impact amount of the second behavior data slice on the first behavior data slice.

[0073] According to the above implementation, the attenuation degree of the influence of the second behavior data segment on the first behavior data segment can be determined based on the time distance between the first behavior data segment and the second behavior data segment, and the preset attenuation coefficient. Thus, based on the attenuation degree of the influence, the first feature of the second behavior data segment is attenuated to obtain the feature influence of the second behavior data segment on the first behavior data segment.

[0074] In one embodiment, based on the feature influence of each second behavior data slice on the first behavior data slice, the first feature of the first behavior data slice is feature modified to obtain the second feature of the first behavior data slice, including: summing the feature influence of each second behavior data slice on the first behavior data slice to obtain the total feature influence; summing the total feature influence and the first feature of the first behavior data slice to obtain the second feature of the first behavior data slice.

[0075] Exemplarily, a corresponding weight value is determined for each second behavior data slice, and the weight value of each second behavior data slice is used to perform weighted summation on the feature influence of each second behavior data slice on the first behavior data slice to obtain a total feature influence.

[0076] Exemplarily, based on the ratio between the total feature influence and the first behavior data slice, the weight values ​​of the two are determined, and then the two weight values ​​are used to sum the feature influence and the first feature of the first behavior data slice to obtain the second feature of the first behavior data slice.

[0077] According to the above implementation, the feature influence of each second behavior data slice on the first behavior data slice is summed to obtain the total feature influence. Then, the total feature influence and the first feature of the first behavior data slice are summed to obtain the second feature of the first behavior data slice.

[0078] In one embodiment, the second feature includes the frequency of occurrence of each dimensional object under each dimensional category in the behavior data slice, and based on the second feature of each behavior data slice, determining whether the first device is an application-activated cheating device includes: determining the third feature of the third behavior data slice under each dimensional category based on the frequency of occurrence of each dimensional object under each dimensional category in the third behavior data slice; inputting the third feature of the third behavior data slice under each dimensional category into the cheating discrimination model to obtain a discrimination result output by the cheating discrimination model, wherein the discrimination result is used to characterize whether the first device is an application-activated cheating device.

[0079] Exemplarily, a dimension category may be a category corresponding to a single dimension, or a combined category corresponding to multiple dimensions. Each category may include one or more dimensions. Each dimension may include one or more dimension objects. For example, for the dimension of surname, it may include multiple surnames such as Chen, Wang, and Li. These specific surnames are dimension objects.

[0080] For example, the dimensions may include device dimension, region dimension and channel dimension. Figure 2 As shown, multiple dimension combinations are exemplified.

[0081] like Figure 2 As shown in the figure, you can select some dimensions and then combine them. For each dimension combination, you can calculate the features of the behavior data to obtain various types of features, such as aggregation features, bucket features, and count features. These behavior data can be divided according to time windows, and the behavior data in a fixed window can be calculated. Then, the calculated features are output to the model.

[0082] Exemplarily, the third feature may include a proportion feature and the above-mentioned deduplication count feature. The proportion feature may include a concentration feature and the like.

[0083] Exemplarily, the third behavior data slice may be a slice in the activation behavior data. For example, by using the first feature of each behavior data slice, a target slice is determined in the activation behavior data, and the target slice is the third behavior data slice.

[0084] Exemplarily, the cheating discrimination model may adopt a general tree model, such as a gradient boosting tree.

[0085] According to the above implementation, the frequency of occurrence of each dimension object under each dimension in the activation behavior data in the specified slice is statistically analyzed to obtain the third feature. The third feature is then input into the cheating discrimination model to obtain the discrimination result output by the cheating discrimination model. Thus, a more accurate discrimination result can be further obtained.

[0086] In one embodiment, the third feature includes a proportion feature, and the third feature of the third behavior data slice under each dimensional category is determined based on the occurrence frequency of each dimensional object in the third behavior data slice under each dimensional category, including: determining at least one target occurrence frequency that meets preset conditions among the occurrence frequencies of each dimensional object in the third behavior data slice under the first dimensional category; summing the at least one target occurrence frequency to obtain a first value; summing the occurrence frequencies of all dimensional objects under the first dimensional category in the third behavior data slice to obtain a second value; and determining the proportion feature of the third behavior data slice under the first dimensional category based on the ratio of the first value to the second value.

[0087] It can be understood that, for any dimension category, the above implementation method can be used to calculate the proportion characteristics of the third behavior data fragment under the dimension category.

[0088] It can be understood that the first dimension category is any dimension category among the optional dimension categories.

[0089] It is understandable that the preset condition can be set as needed. For example, the preset condition is that the dimension object is a preset object. Exemplarily, the preset object can be the traffic of the network access point as a preset operator.

[0090] According to the above implementation, the occurrence frequency of at least one target that meets the preset conditions is first determined; the occurrence frequency of the at least one target is summed to obtain a first value, and the occurrence frequency of all dimensional objects under the first dimensional category in the third behavior data slice is summed to obtain a second value. Then, based on the ratio of the first value to the second value, the proportion characteristics of the third behavior data slice under the first dimensional category can be accurately determined.

[0091] In one embodiment, the proportion feature includes a concentration feature, and the preset condition is that the target occurrence frequency is ranked in the top N positions of the occurrence frequencies of all dimensional objects under the first dimensional category, where N is a positive integer greater than 1.

[0092] It can be understood that the value of N is less than or equal to the total number of all dimensional objects under the first dimensional category.

[0093] According to the above implementation, by setting the pre-selection condition as the top N digits of the target occurrence frequency ranking among the occurrence frequencies of all dimensional objects under the first dimensional category, and combining it with the calculation process of the above-mentioned proportion feature, the concentration feature of the third behavior data segment under the first dimensional category can be obtained.

[0094] Figure 31 is a flow chart of an anti-cheating processing method of an embodiment of the present disclosure. The method specifically includes three stages. Among them, the second stage is applied to activate the cheating device judgment system. The system includes a data processing center, a feature calculation module, a pre-processing module, and a cheating prediction / model training module.

[0095] The first stage is to collect data. When the device downloads the application, the application carries a data collection module. This module is the source of data for the entire system. Its main function is to collect and report the activation information and device information triggered by the user within a short period of time after the device downloads the APP and opens the APP. This module will standardize the device information that can be collected and transfer it to the activation cheating device judgment system for feature calculation.

[0096] In the second stage, the model is trained or the data is judged for cheating. The feature calculation module in the cheating device judgment system is activated to realize feature calculation of streaming or offline data transmitted by the data acquisition module within a specified time window. This module performs feature statistics on the data contained in any dimension or any combination of multiple dimensions within a time window of a specified length. The categories of features can include concentration, proportion, deduplication count, etc. After the features are accumulated, the data is stored on the disk and provided to the pre-processing module for data processing.

[0097] Among them, the feature calculation module can be completed based on themis-mr framework, supporting data warehouse tables and HDFS files as static data sources, or it can be calculated based on the flink streaming framework, supporting real-time calculation of the front window of streaming data.

[0098] The pre-processing module in the activation cheating device judgment system may also be referred to as a sample enhancement module.

[0099] This module is a pre-dependent module for model training / data prediction. It processes the data from the feature calculation module in breadth and depth. In the model training phase, the data from the disk is oversampled, undersampled, and / or noise data is introduced in a specific proportion to obtain the training data set. In the data prediction phase, the above-mentioned oversampling, undersampling, and noise data introduction processing are not required.

[0100] Among them, feature breadth processing can include:

[0101] 1. In the scenario of channel activation fraud identification, introduce channel-related characteristics and make one-hot mapping features. For example, the characteristics of the channel can include paid channels, free channels, channel charging methods, etc. Then, one-hot map these features to obtain one-hot mapping features. For example, suppose there are three types of channels, including free, paid, and pre-installed. After one-hot mapping these three types, we can get free channels 001, paid channels 010, and pre-installed channels 100. Finally, match the mapping features with specific sample data.

[0102] 2. Extract features from the inactive behavior data (if any) in the current APP and normalize the features. For example, map each feature to the range of 0-1. This can speed up the speed and accuracy of model calculation in subsequent model training or model prediction.

[0103] 3. Obtain the source tracing information of the APP downloaded by the user. The source tracing information is the source channel of the user's download.

[0104] Among them, feature depth processing can include:

[0105] 1. Map character type features to floating point type features to facilitate model calculation.

[0106] 2. The calculation time window for filtering features, that is, the time length of the above-mentioned behavior data segmentation.

[0107] 3. Clean or delete abnormal or useless values ​​in the features.

[0108] 4. Introducing noise into the features can prevent overfitting during subsequent model training.

[0109] During the model training process, the sample data can also be sampled and processed as follows: Control the ratio of positive and negative samples within a reasonable range. Generally, if the difference between the positive and negative sample ratios exceeds 10 times, the problem of uneven sample distribution will be introduced. This example controls the under-sampling ratio of negative samples to keep the ratio of positive and negative samples within a reasonable range, which can prevent data distribution deviation and avoid gradient disappearance.

[0110] The cheating prediction / model training module is the final decision-making module of the system. Model training is performed in the testing phase, and cheating data identification is performed in the actual application phase.

[0111] The third stage is anti-cheating processing. The growth delivery model side of the APP receives the cheating identification results and automatically adjusts the growth delivery model. The channel cooperation business side of the APP receives the cheating identification results and deducts the settlement amount from the cheating channel to prevent capital loss. The subsequent business link business side of the APP receives the cheating identification results to prevent cheating devices from cheating in subsequent searches, feeds, promotional activities, etc.

[0112] According to the scheme provided by the above implementation method, it can be specifically applied to the identification of cheating devices such as machine brushing, batch, and virtual in the growth investment scenario of APP, and combat the illegal behavior of the black and gray industries in the society to illegally obtain the growth investment funds of APP and the destruction of the APP mobile ecology. This technical solution is relatively general and can be widely used to identify cheating devices in any APP activation and invitation scenario, such as other business fields based on mobile terminal devices such as other search, social, shopping, travel, Q&A, local life, video, etc.

[0113] Figure 4 It is a structural block diagram of a device detection apparatus according to an embodiment of the present disclosure.

[0114] like Figure 4 As shown, the device detection device may include:

[0115] The data acquisition module 410 is used to acquire activation behavior data of the application in the first device;

[0116] A feature calculation module 420, configured to perform feature calculation on each behavior data slice in the activation behavior data to obtain a first feature of each behavior data slice;

[0117] A feature correction module 430, configured to correct the first feature of each of the behavior data slices based on the time distance between each of the behavior data slices to obtain a second feature of each of the behavior data slices;

[0118] The cheating detection module 440 is used to determine whether the first device is an application activation cheating device based on the second feature of each of the behavior data slices.

[0119] Figure 5 It is a structural block diagram of a device detection apparatus according to another embodiment of the present disclosure.

[0120] like Figure 4 and Figure 5 As shown, Figure 5 The data acquisition module 510, the feature calculation module 520, the feature correction module 530 and the cheating detection module 540 are respectively Figure 4The structures and functions of the data acquisition module 410, feature calculation module 420, feature correction module 430 and cheating detection module 440 are the same and will not be described in detail here.

[0121] In one implementation, the feature correction module 530 includes:

[0122] A feature attenuation unit 531 is used to perform feature attenuation on the first feature of each second behavior data slice based on the time distance between the first behavior data slice and each second behavior data slice, so as to obtain the feature influence amount of each second behavior data slice on the first behavior data slice;

[0123] The feature correction unit 532 is used to correct the first feature of the first behavior data slice based on the feature influence of each of the second behavior data slices on the first behavior data slice to obtain the second feature of the first behavior data slice.

[0124] In one implementation, the feature attenuation unit 531 is specifically configured to:

[0125] Determine, based on the time distance between the first behavior data segment and the second behavior data segment, and a preset attenuation coefficient, the attenuation degree of the influence of the second behavior data segment on the first behavior data segment;

[0126] Based on the influence attenuation degree, the first feature of the second behavior data slice is subjected to feature attenuation to obtain the feature influence amount of the second behavior data slice on the first behavior data slice.

[0127] In one implementation, the feature correction unit 532 is specifically configured to:

[0128] Sum the feature influence of each of the second behavior data slices on the first behavior data slice to obtain a total feature influence;

[0129] The total feature influence amount and the first feature of the first behavior data slice are summed to obtain the second feature of the first behavior data slice.

[0130] In one implementation, the second feature includes the occurrence frequency of each dimensional object under each dimensional category in the behavior data shard, and the cheating detection module 540 includes:

[0131] The feature statistics unit 541 is used to determine the third feature of the third behavior data slice under each dimension category based on the occurrence frequency of each dimension object under each dimension category in the third behavior data slice;

[0132] The model processing unit 542 is used to input the third feature of the third behavior data segment under each dimension category into the cheating discrimination model to obtain a discrimination result output by the cheating discrimination model, wherein the discrimination result is used to indicate whether the first device is an application-activated cheating device.

[0133] In one implementation, the third feature includes a proportion feature, and the feature statistics unit 541 is specifically configured to:

[0134] Determine at least one target occurrence frequency that meets a preset condition among the occurrence frequencies of each dimensional object in the third behavior data slice under the first dimensional category;

[0135] Summing the occurrence frequency of the at least one target to obtain a first value;

[0136] Sum the occurrence frequencies of all dimensional objects under the first dimensional category in the third line of data shards to obtain a second value;

[0137] Based on the ratio of the first value to the second value, the proportion feature of the third behavior data slice under the first dimension category is determined.

[0138] In one embodiment, the proportion feature includes a concentration feature, and the preset condition is that the target occurrence frequency is ranked in the top N positions among the occurrence frequencies of all dimensional objects under the first dimensional category, where N is a positive integer greater than 1.

[0139] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present disclosure, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0140] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0141] According to an embodiment of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium and a computer program product. The electronic device may be a server, and the server may also include any power supply control device in the embodiment of the present disclosure.

[0142] Figure 6A schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0143] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0144] Multiple components in the electronic device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0145] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 601 performs the various methods and processes described above, such as a device detection method. For example, in some embodiments, a device detection method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps in the device detection method described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform a device detection method in any other appropriate manner (e.g., by means of firmware).

[0146] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0148] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0149] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0150] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0151] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0152] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0153] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A device detection method, comprising: Acquire activation behavior data of an application in the first device; Performing feature calculation on each behavior data slice in the activation behavior data to obtain a first feature of each behavior data slice; Based on the time distance between each of the behavior data slices, feature correction is performed on the first feature of each of the behavior data slices to obtain the second feature of each of the behavior data slices; Based on the second feature of each of the behavior data slices, determine whether the first device is an application activation cheating device.

2. The method according to claim 1, wherein: The first feature of each behavior data slice is modified based on the time distance between each behavior data slice to obtain the second feature of each behavior data slice, including: Based on the time distance between the first behavior data slice and each second behavior data slice, feature attenuation is performed on the first feature of each second behavior data slice to obtain the feature influence amount of each second behavior data slice on the first behavior data slice; Based on the amount of influence of each of the second behavior data slices on the characteristics of the first behavior data slice, the first characteristics of the first behavior data slice are modified to obtain the second characteristics of the first behavior data slice.

3. The method according to claim 2, wherein: performing feature attenuation on the first feature of each of the second behavior data slices based on the time distance between the first behavior data slice and each of the second behavior data slices, Obtaining the characteristic influence of each of the second behavior data slices on the first behavior data slice includes: Determine, based on the time distance between the first behavior data segment and the second behavior data segment, and a preset attenuation coefficient, the attenuation degree of the influence of the second behavior data segment on the first behavior data segment; Based on the influence attenuation degree, the first feature of the second behavior data slice is subjected to feature attenuation to obtain the feature influence amount of the second behavior data slice on the first behavior data slice.

4. The method according to claim 2, wherein: The step of modifying the first feature of the first behavior data slice based on the feature influence of each of the second behavior data slices on the first behavior data slice to obtain the second feature of the first behavior data slice includes: Sum the feature influence of each of the second behavior data slices on the first behavior data slice to obtain a total feature influence; The total feature influence amount and the first feature of the first behavior data slice are summed to obtain the second feature of the first behavior data slice.

5. The method according to claim 1, wherein: The second feature includes the occurrence frequency of each dimension object under each dimension category in the behavior data slice, and determining whether the first device is an application activation cheating device based on the second feature of each behavior data slice includes: Determine, based on the occurrence frequency of each of the dimensional objects in the third behavior data slice under each of the dimensional categories, the third feature of the third behavior data slice under each of the dimensional categories; The third feature of the third behavior data segment under each dimension category is input into a cheating discrimination model to obtain a discrimination result output by the cheating discrimination model, wherein the discrimination result is used to characterize whether the first device is an application-activated cheating device.

6. The method according to claim 5, wherein: The third feature includes a proportion feature, and the third feature of the third behavior data slice under each dimension category is determined based on the occurrence frequency of each dimension object under each dimension category in the third behavior data slice, including: Determine at least one target occurrence frequency that meets a preset condition among the occurrence frequencies of each dimensional object in the third behavior data slice under the first dimensional category; Summing the occurrence frequency of the at least one target to obtain a first value; Sum the occurrence frequencies of all dimensional objects under the first dimensional category in the third line of data shards to obtain a second value; Based on the ratio of the first value to the second value, the proportion feature of the third behavior data slice under the first dimension category is determined.

7. The method according to claim 6, wherein: The proportion feature includes a concentration feature, and the preset condition is that the target occurrence frequency is ranked in the top N positions among the occurrence frequencies of all dimensional objects under the first dimensional category, where N is a positive integer greater than 1.

8. A device detection apparatus, comprising: A data acquisition module, used to acquire activation behavior data of an application in the first device; A feature calculation module, used for performing feature calculation on each behavior data slice in the activation behavior data to obtain a first feature of each behavior data slice; A feature correction module, used for correcting the first feature of each of the behavior data slices based on the time distance between each of the behavior data slices to obtain a second feature of each of the behavior data slices; The cheating detection module is used to determine whether the first device is an application-activated cheating device based on the second feature of each of the behavior data slices.

9. The device according to claim 8, wherein: The feature correction module comprises: a feature attenuation unit, configured to perform feature attenuation on the first feature of each second behavior data slice based on the time distance between the first behavior data slice and each second behavior data slice, and obtain a feature influence amount of each second behavior data slice on the first behavior data slice; The feature correction unit is used to correct the first feature of the first behavior data slice based on the feature influence of each of the second behavior data slices on the first behavior data slice to obtain the second feature of the first behavior data slice.

10. The device according to claim 9, wherein: The characteristic attenuation unit is specifically used for: Determine, based on the time distance between the first behavior data segment and the second behavior data segment, and a preset attenuation coefficient, the attenuation degree of the influence of the second behavior data segment on the first behavior data segment; Based on the influence attenuation degree, the first feature of the second behavior data slice is subjected to feature attenuation to obtain the feature influence amount of the second behavior data slice on the first behavior data slice.

11. The device according to claim 9, wherein: The feature correction unit is specifically used for: Sum the feature influence of each of the second behavior data slices on the first behavior data slice to obtain a total feature influence; The total feature influence amount and the first feature of the first behavior data slice are summed to obtain the second feature of the first behavior data slice.

12. The device according to claim 8, wherein: The second feature includes the frequency of occurrence of each dimensional object under each dimensional category in the behavior data shard, and the cheating detection module includes: A feature statistics unit, configured to determine a third feature of the third behavior data slice under each of the dimensional categories based on the occurrence frequency of each of the dimensional objects under each of the dimensional categories in the third behavior data slice; The model processing unit is used to input the third feature of the third behavior data segment under each dimension category into the cheating discrimination model to obtain a discrimination result output by the cheating discrimination model, wherein the discrimination result is used to characterize whether the first device is an application-activated cheating device.

13. The device according to claim 12, wherein: The third feature includes a proportion feature, and the feature statistics unit is specifically used to: Determine at least one target occurrence frequency that meets a preset condition among the occurrence frequencies of each dimensional object in the third behavior data slice under the first dimensional category; Summing the occurrence frequency of the at least one target to obtain a first value; Sum the occurrence frequencies of all dimensional objects under the first dimensional category in the third line of data shards to obtain a second value; Based on the ratio of the first value to the second value, the proportion feature of the third behavior data slice under the first dimension category is determined.

14. The device according to claim 13, wherein: The proportion feature includes a concentration feature, and the preset condition is that the target occurrence frequency is ranked in the top N positions among the occurrence frequencies of all dimensional objects under the first dimensional category, where N is a positive integer greater than 1.

15. An electronic device, comprising: at least one central processing unit; as well as A memory in communication with the at least one central processing unit; wherein, The memory stores instructions that can be executed by the at least one central processor, and the instructions are executed by the at least one central processor to enable the at least one central processor to perform the method according to any one of claims 1 to 7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.

17. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Determination method of cheat user and equipment thereof

    CN106296203A

  • Method and apparatus for identifying device cheating in APP promotion

    CN107153971A

  • Equipment activation method and device, equipment and storage medium

    CN118138313A

  • Clustering for Detection of Anomalous Behavior and Insider Threat

    US20180359270A1

  • Cheat detection by comparing mouse input and frame buffer

    US20230241512A1