User identification method, device, electronic device and non-volatile storage medium

Abstract

The present application discloses a user identification method, device, electronic device and non-volatile storage medium. The method includes: determining the target attribute based on the information gain value of each type of attribute data in the target data, wherein the information gain value is used to characterize the degree of influence of each attribute on the judgment of the user account category, and the user account category includes: live streaming users and non-live streaming users; obtaining the target association relationship by calculating the posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category; determining the user account category of each user account in the target data based on the target association relationship. The present application solves the technical problem of low efficiency in identifying live streaming users because the related technology can only discover a small number of live streaming unicast streaming users through manual auditing of call records, on-site inspections, etc., and cannot accurately identify all streaming users.

Description

Technical Field

[0001] The present application relates to the technical field of network television live broadcast, and in particular, to a user identification method, device, electronic device and non-volatile storage medium. Background Art

[0002] With the rapid development of the live broadcast industry, the competition among video contents is becoming increasingly fierce, and the number of live broadcast platform traffic thefts is increasing. Since live broadcast content is played to the public, if the traffic theft users enter the live broadcast room through ordinary users to watch, and then play the content they watch through the playback software of other platforms, it will cause serious loss of content on the live broadcast platform, reduce the value of video content, and lose user flow.

[0003] The current live streaming stealing user identification technology is that when users complain that the IPTV (interactive network television) service is stuck or cannot be watched, the platform maintenance personnel manually audit the user account status and historical call records, and find abnormal behavior such as requesting multiple streams at the same time, and infer that the user is a live streaming stealing user. Or when the installation and maintenance or customer manager visits the site for troubleshooting, they find that the user's IPTV service networking is not the telecommunications standard networking method, but the IPTV service stream is obtained by connecting to a private server, and infer that the user is a live streaming stealing user. The relevant technology can only detect a small number of live unicast stealing users through manual methods such as manual auditing of call records and on-site inspections, and cannot accurately identify all stealing users. There are problems such as low efficiency in identifying live streaming stealing users.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a user identification method, device, electronic device and non-volatile storage medium to at least solve the technical problem of low efficiency in identifying live unicast traffic theft users, as the related technology can only discover a small number of live unicast traffic theft users through manual audit of call records, on-site inspections, etc., but cannot accurately identify all the traffic theft users.

[0006] According to one aspect of an embodiment of the present application, a user identification method is provided, including: discretizing live broadcast service data to obtain target data; determining target attributes based on information gain values ​​of various attribute data in the target data, wherein the information gain value is used to characterize the degree of influence of each attribute on determining a user account category, and the user account categories include: live broadcast stream stealing users and non-live broadcast stream stealing users; obtaining a target association relationship by calculating a posterior probability value that a user account is of each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category; and determining the user account category of each user account in the target data based on the target association relationship.

[0007] Optionally, determining the target attribute based on the information gain value of each type of attribute data in the target data includes: determining the information gain value of each type of attribute data in the target data; and when the information gain value is greater than a preset gain threshold, determining the attribute corresponding to the information gain value as the target attribute.

[0008] Optionally, determining the information gain value of each type of attribute data in the target data includes: calculating a first information entropy of the target data, wherein the first information entropy is used to characterize the amount of information provided by the target data for determining the category of the user account; respectively calculating a second information entropy of each type of attribute data in the target data, wherein the second information entropy is used to characterize the amount of information provided by each type of attribute data in the target data for determining the category of the user account; and performing calculations based on the first information entropy and the second information entropy to obtain information gain values ​​corresponding to each type of attribute in the target data.

[0009] Optionally, obtaining the target association relationship by calculating the posterior probability value of the user account being each user account category includes: obtaining training sample data based on the target attribute, wherein the training sample data includes multiple sample data of multiple user accounts, and each sample data includes the target attribute and the user account category corresponding to the attribute value of the target attribute; based on the training sample data, training the classification model by calculating the posterior probability value of the user account being each user account category under the condition that each attribute value in the attribute value sequence of the target attribute is in different states, to obtain the target association relationship, wherein the attribute value sequence includes all the target attributes, and the state of the attribute value of each target attribute includes: yes and no.

[0010] Optionally, calculating the posterior probability value to train the classification model includes: calculating the prior probability value of each user account category based on the training sample data; calculating the likelihood probability value corresponding to each attribute value in the attribute value sequence of the target attribute under different states under the condition that the user account is each user account category; calculating based on the prior probability value and the likelihood probability value to obtain the posterior probability value; predicting the user account category of the user account based on the posterior probability value, and comparing it with the user account category corresponding to the user account pre-stored in the training sample data.

[0011] Optionally, calculating the likelihood probability value corresponding to the attribute value sequence of the target attribute in each case under the condition that the user account is assumed to be each user account category includes: calculating the candidate probability value corresponding to the attribute value of each target attribute in each state under the condition that the user account is assumed to be each user account category; and calculating the likelihood probability value based on the candidate probability value.

[0012] Optionally, discretizing the live broadcast service data to obtain target data includes: encoding and integrating the attribute values ​​of each attribute in the live broadcast service data according to a preset encoding rule, wherein the states of the attribute values ​​of each attribute after encoding and integration include: yes and no.

[0013] According to another aspect of an embodiment of the present application, a user identification device is also provided, including: a preprocessing module, used to discretize live broadcast business data to obtain target data; an attribute determination module, used to determine the target attribute based on the information gain value of each attribute data in the target data, wherein the information gain value is used to characterize the influence of each attribute on the judgment of the user account category, and the user account category includes: live broadcast flow stealing users and non-live broadcast flow stealing users; a probability calculation module, used to obtain a target association relationship by calculating the posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category; a category identification module, used to determine the user account category of each user account in the target data based on the target association relationship.

[0014] According to another aspect of the embodiments of the present application, an electronic device is provided. The electronic device includes a processor, and the processor is used to run a program, wherein the user identification method is executed when the program is run.

[0015] According to another aspect of the embodiments of the present application, a non-volatile storage medium is provided, the non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the user identification method by running the computer program.

[0016] In the embodiment of the present application, the live broadcast service data is discretized to obtain target data; the target attribute is determined according to the information gain value of each attribute data in the target data, wherein the information gain value is used to characterize the influence degree of each attribute on the determination of the user account category, and the user account category includes: live broadcast flow theft user and non-live broadcast flow theft user; the target association relationship is obtained by calculating the posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category; according to the target association relationship, the user account category of each user account in the target data is determined in a manner that the user behavior characteristics represented by the live broadcast service data of the user are associated with the live broadcast flow theft behavior, and whether the user is a live broadcast flow theft user is determined according to the probability relationship between the user behavior characteristics and the live broadcast flow theft behavior, thereby achieving the purpose of improving the efficiency of identifying live broadcast flow theft users, and further solving the technical problem of low efficiency of identifying live broadcast flow theft users caused by the fact that the related technology can only discover a small number of live broadcast unicast flow theft users through manual audit of call records, on-site inspections, etc., and cannot accurately identify all the flow theft users. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 It is a hardware structure block diagram of a computer terminal (or electronic device) for implementing a method for user identification provided in an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of a user identification method flow provided according to an embodiment of the present application;

[0020] Figure 3 It is a flowchart of a method for determining target attributes provided in an embodiment of the present application;

[0021] Figure 4 It is a flowchart of a method for calculating a posterior probability value to obtain a target association relationship according to an embodiment of the present application;

[0022] Figure 5 It is a structural schematic diagram of a user identification device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] In the related technologies, only after the user complains, can the live unicast traffic theft users be discovered through manual audit of call records, on-site inspections, etc., which lacks active discovery means and has low manual identification accuracy. At the same time, the live multicast traffic theft users cannot be discovered by this method, and the traffic theft users cannot be completely discovered, resulting in the problem of inaccurate identification of live broadcast traffic theft users;

[0026] In addition, the related technology can only detect a small number of live unicast stealing users, and cannot accurately identify all stealing users. When blocking stealing, it can only implement point-to-point blocking for unicast stealing users, and cannot implement comprehensive blocking for multicast stealing and live unicast stealing users. There is a problem of incomplete blocking of live stealing users;

[0027] In order to solve the above problems, relevant solutions are provided in the embodiments of the present application, which are described in detail below.

[0028] According to an embodiment of the present application, a method embodiment of user identification is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0029] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1FIG. 1 shows a hardware structure block diagram of a computer terminal (or electronic device) for implementing a user identification method. Figure 1 As shown, the computer terminal 10 (or electronic device 10) may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0030] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or electronic device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0031] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the user identification method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned user identification method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0032] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or electronic device).

[0034] In the above operating environment, the embodiment of the present application provides a user identification method. Figure 2 is a schematic diagram of a user identification method flow according to an embodiment of the present application, such as Figure 2 As shown, the method comprises the following steps:

[0035] Step S202, discretizing the live broadcast service data to obtain target data;

[0036] The above live broadcast business data are BO domain data such as platform logs, network status, user playback behavior, etc. of all users after integration and cleaning.

[0037] In some embodiments of the present application, the live broadcast service data is discretized to obtain the target data, including the following steps: according to a preset coding rule, the attribute value of each attribute in the live broadcast service data is encoded and integrated, wherein the status of the attribute value of each attribute after encoding and integration includes: yes and no.

[0038] Specifically, according to the IPTV platform data and network equipment status, data integration and cleaning are performed on user data and service attributes (number of URLs requested by users, type of requested programs, regularity of power on and off time, number of CDNs served by users, platform network attributes such as downstream bandwidth, user groups, regions and other user data) that have live unicast RTSP requests or live multicast IGMP requests at the same time point. The amount of data is reduced by deleting irrelevant or redundant attributes to obtain integrated data (i.e., the above-mentioned live service data).

[0039] The continuous variables in the attributes of the integrated data (i.e. the live broadcast business data mentioned above) are discretized, and the discrete variables and text attributes are encoded to achieve category integration and obtain the target data.

[0040] Step S204, determining the target attribute according to the information gain value of each attribute data in the target data, wherein the information gain value is used to characterize the influence degree of each attribute on determining the user account category, and the user account category includes: live streaming stealing users and non-live streaming stealing users;

[0041] In some embodiments of the present application, determining the target attribute based on the information gain value of each type of attribute data in the target data includes the following steps: determining the information gain value of each type of attribute data in the target data; when the information gain value is greater than a preset gain threshold, determining that the attribute corresponding to the information gain value is the target attribute.

[0042] In some embodiments of the present application, determining the information gain value of each type of attribute data in the target data includes the following steps: calculating the first information entropy of the target data, wherein the first information entropy is used to characterize the amount of information provided by the target data for determining the category of the user account; respectively calculating the second information entropy of each type of attribute data in the target data, wherein the second information entropy is used to characterize the amount of information provided by each type of attribute data in the target data for determining the category of the user account; performing calculations based on the first information entropy and the second information entropy to obtain the information gain value corresponding to each type of attribute in the target data.

[0043] Figure 3 is a flow chart of a method for determining target attributes according to an embodiment of the present application, such as Figure 3 As shown, the method comprises the following steps:

[0044] Step S302, calculate the total information entropy I(s1, s2…s m )(i.e. the first information entropy mentioned above);

[0045]

[0046] Among them, s1, s2…s m is the m sample data in the above target data, m is the number of samples in the target data, P i Any sample belongs to C i The probability that C i It includes C1 and C2, C1 indicates that the user account corresponding to the sample data is a live streaming stealing user, and C2 indicates that the user account corresponding to the sample data is not a live streaming stealing user.

[0047] Step S304, calculating the information entropy E(A) of each attribute (ie, the second information entropy mentioned above);

[0048]

[0049] in, k represents the number of attributes in the target data, s1j ,s 2j …s mj represents each attribute in each sample data in the target data, s represents the total sample set, P ij Indicates that the category in the subset is C i The probability of a sample of .

[0050] Step S306, calculate the information gain value of each attribute Gain(A) = I(s1, s2…s m )-E(A);

[0051] The smaller E(A) is, the larger the value of Gain(A) is, which means that the more information the test attribute A provides for classification, and the smaller the uncertainty of classification after selecting A.

[0052] Step S308, determining target attributes;

[0053] As an optional implementation, a preset number of attributes with maximum information gain values ​​may be selected as target attributes.

[0054] For example, the top 10 dimensions with the largest information gain are selected as target attributes, as shown in the following table.

[0055]

[0056]

[0057] Step S206, obtaining a target association relationship by calculating a posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category;

[0058] In some embodiments of the present application, obtaining a target association relationship by calculating a posterior probability value of a user account for each user account category includes the following steps: obtaining training sample data based on a target attribute, wherein the training sample data includes multiple sample data of multiple user accounts, and each sample data includes a target attribute and a user account category corresponding to an attribute value of the target attribute; based on the training sample data, training a classification model by calculating a posterior probability value of a user account for each user account category under the condition that each attribute value in an attribute value sequence of the target attribute is in different states, and obtaining a target association relationship, wherein the attribute value sequence includes all target attributes, and the state of the attribute value of each target attribute includes: yes and no.

[0059] In some embodiments of the present application, calculating the posterior probability value to train the classification model includes the following steps: calculating the prior probability value of each user account category based on the training sample data; calculating the likelihood probability value corresponding to each attribute value in the attribute value sequence of the target attribute under the condition that the user account is each user account category; calculating based on the prior probability value and the likelihood probability value to obtain the posterior probability value; predicting the user account category of the user account based on the posterior probability value, and comparing it with the user account category corresponding to the user account pre-stored in the training sample data.

[0060] In some embodiments of the present application, calculating the likelihood probability value corresponding to the attribute value sequence of the target attribute in each case under the condition that the user account is assumed to be each user account category includes the following steps: calculating the candidate probability value corresponding to the attribute value of each target attribute in each state under the condition that the user account is assumed to be each user account category; and calculating the likelihood probability value based on the candidate probability value.

[0061] Figure 4 FIG. 1 is a flow chart of a method for calculating a posterior probability value to obtain a target association relationship according to an embodiment of the present application. Figure 4 As shown, the method comprises the following steps:

[0062] Step S402, estimating the probability of each category appearing (i.e. the above-mentioned prior probability value);

[0063] Specifically, based on the training sample data, the prior probability P(C i )(i=1,2, in this embodiment, C i It includes C1 and C2, C1 indicates that the user account corresponding to the sample data is a live streaming stealing user, and C2 indicates that the user account corresponding to the sample data is not a live streaming stealing user.

[0064] Step S404, estimating the probability of occurrence of each attribute value under each category condition (ie, the candidate probability value mentioned above);

[0065] Specifically, P(x k |C i )(k=1,2...n), calculate the probability P(x1|C) from the training sample data i ), P(x2|C i )…P(x n |C i ), after discretization encoding, all attributes are discrete data, P(x k |C i ) is the attribute A in the training sample data k The value of x k The number of samples divided by the number of samples in the training tuple C iThe number of samples of the class.

[0066] Step S406, calculating the probability of each attribute combination belonging to each category (ie, the likelihood probability value mentioned above);

[0067] Specifically, the calculation condition C i The likelihood probability P(X|C i );

[0068]

[0069] Among them, X is the above attribute value sequence.

[0070] Step S408: Select the class with the largest posterior probability (ie, the user account category) as the inference result output for the data.

[0071] Specifically, under condition X, the naive Bayes classification method is used to predict that X belongs to the class with the highest posterior probability (ie, the above-mentioned user account category), the prediction result is output, and compared with the user account category corresponding to the user account pre-stored in the training sample data.

[0072] The following is an example to illustrate the method in step S402 to step S408.

[0073] The training data (i.e. the above-mentioned training sample data) formed after selecting 100 sample data for data preprocessing and target attribute selection is shown in the following table.

[0074]

[0075] The test users were used to conduct classification and recognition tests. The test users’ data are shown in the following table.

[0076]

[0077]

[0078] The meanings of the target attributes of the test user are shown in the following table;

[0079]

[0080] The test user is a hotel user in an urban area. The user has been online for more than 12 hours and has regular power on and off behavior every day. At the same time, the user requests more than 2 RTSP requests. According to manual experience, normal users will not be online for a long time, and there is no fixed pattern for daily power on and off. In addition, the number of requests made by this user at the same time is too large, so it can be judged that this user is an abnormal user, that is, this user is a live streaming user.

[0081] The steps for making judgments using this application scheme include:

[0082] Step 1: Train the naive Bayes algorithm based on 100 household training samples and use the sample attributes to calculate the class prior probability value P(C i ), and the conditional probability P(x k |C i );

[0083] For example, the class prior probabilities obtained based on the training samples are as follows: the probability of live streaming stealing users in the training samples is 0.357, and the probability of non-live streaming stealing users is 0.643.

[0084] P(C1)=0.357,P(C2)=0.643

[0085] Step 2: Take the probability obtained in step 1 as prior knowledge and calculate the attribute value sequence X of each attribute of the test user under condition C. i The likelihood probability P(X|C i );

[0086] P(X|C=1)=0.0003535, P(X|C=0)=0.00001848

[0087] Step 3, based on the conditional independence assumption, we only need to find the posterior probability value P(X|C i )P(C i ) maximized user account category;

[0088] P(X|C1)P(C1)=0.0003535×0.357=0.0001262

[0089] P(X|C2)P(C2)=0.00001848×0.643=0.00001188

[0090] Therefore, for the test user, the posterior probability of belonging to category 1, 0.0001262, is greater than the posterior probability of belonging to category 0, and the user account category of the user is predicted to be a live streaming stealing user.

[0091] Step S208, determining the user account category of each user account in the target data according to the target association relationship.

[0092] Specifically, the user's original data is processed and input into the target association relationship obtained in the above steps, and a list of live streaming stealing users for a period of time is output.

[0093] For example, the meaning of the fields of the output live streaming stealing user list are shown in the following table. According to the user list, users can be accurately located to assist staff in on-site inspection and implementation of stealing blocking.

[0094]

[0095] Through the above steps, by correlating the user behavior characteristics represented by the user's live broadcast business data with the live broadcast theft behavior, it is determined whether the user is a live broadcast theft user based on the probabilistic relationship between the user behavior characteristics and the live broadcast theft behavior, thereby achieving the purpose of improving the efficiency of identifying live broadcast theft users, and further solving the technical problem of low efficiency in identifying live broadcast theft users caused by the fact that related technologies can only discover a small number of live broadcast unicast theft users through manual auditing of call records, on-site inspections, etc., and cannot accurately identify all the theft users.

[0096] According to an embodiment of the present application, an embodiment of a user identification device is also provided. Figure 5 is a schematic diagram of the structure of a user identification device provided according to an embodiment of the present application. Figure 5 As shown, the device comprises:

[0097] The preprocessing module 50 is used to discretize the live broadcast service data to obtain target data;

[0098] The attribute determination module 52 is used to determine the target attribute according to the information gain value of each attribute data in the target data, wherein the information gain value is used to characterize the influence degree of each attribute on the determination of the user account category, and the user account category includes: live streaming stealing users and non-live streaming stealing users;

[0099] The probability calculation module 54 is used to obtain the target association relationship by calculating the posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category;

[0100] The category identification module 56 is used to determine the user account category of each user account in the target data according to the target association relationship.

[0101] It should be noted that the various modules in the above-mentioned user identification device can be program modules (for example, a set of program instructions that implement a certain specific function) or hardware modules. For the latter, it can be expressed in the following forms, but is not limited to this: the expression form of each of the above-mentioned modules is a processor, or the functions of each of the above-mentioned modules are implemented by a processor.

[0102] It should be noted that the user identification device provided in this embodiment can be used to perform Figure 2 The user identification method shown, therefore, the relevant explanations and instructions on the above user identification method are also applicable to the embodiments of the present application and will not be repeated here.

[0103] The embodiment of the present application also provides a non-volatile storage medium, the non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the following user identification method by running the computer program: discretizing the live broadcast service data to obtain target data; determining the target attribute based on the information gain value of each attribute data in the target data, wherein the information gain value is used to characterize the influence of each attribute on the judgment of the user account category, and the user account category includes: live broadcast stream stealing users and non-live broadcast stream stealing users; obtaining the target association relationship by calculating the posterior probability value of the user account being each user account category, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category; determining the user account category of each user account in the target data based on the target association relationship.

[0104] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0105] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0106] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0107] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0108] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0109] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0110] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A user identification method, characterized in that: include: Discretize the live broadcast service data to obtain target data; Determine the target attribute according to the information gain value of each type of attribute data in the target data, wherein the information gain value is used to characterize the influence of each attribute on determining the user account category, and the user account category includes: live streaming stealing users and non-live streaming stealing users; The target association relationship is obtained by calculating the posterior probability value of the user account being each of the user account categories, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category, and the target association relationship is obtained by calculating the posterior probability value of the user account being each of the user account categories, including: acquiring training sample data according to the target attribute, wherein the training sample data includes multiple sample data of multiple user accounts, and each sample data includes the target attribute and the user account category corresponding to the attribute value of the target attribute; based on the training sample data, the classification model is trained by calculating the posterior probability value of the user account being each of the user account categories under the condition that each attribute value in the attribute value sequence of the target attribute is in different states, thereby obtaining the target association relationship, wherein the attribute value sequence includes all the target attributes, and the state of each attribute value of the target attribute includes: yes and no; The user account category of each user account in the target data is determined according to the target association relationship.

2. The user identification method according to claim 1, characterized in that: According to the information gain values ​​of various attribute data in the target data, the target attributes are determined to include: Determine the information gain value of each type of attribute data in the target data; When the information gain value is greater than a preset gain threshold, it is determined that the attribute corresponding to the information gain value is the target attribute.

3. The user identification method according to claim 2, characterized in that: Determining the information gain value of each type of attribute data in the target data includes: Calculating a first information entropy of the target data, wherein the first information entropy is used to characterize the amount of information provided by the target data for determining the category of the user account; respectively calculating the second information entropy of each type of attribute data in the target data, wherein the second information entropy is used to characterize the amount of information provided by each type of attribute data in the target data for determining the category of the user account; The information gain value corresponding to each attribute in the target data is obtained by calculating according to the first information entropy and the second information entropy.

4. The user identification method according to claim 1, characterized in that: Calculating the posterior probability value to train the classification model includes: Calculating the prior probability value of each user account category based on the training sample data; Calculating the likelihood probability value corresponding to each attribute value in the attribute value sequence of the target attribute under different states under the condition that the user account is assumed to be each of the user account categories; Calculating according to the prior probability value and the likelihood probability value to obtain the posterior probability value; The user account category of the user account is predicted based on the posterior probability value, and compared with the user account category corresponding to the user account pre-stored in the training sample data.

5. The user identification method according to claim 4, characterized in that: Calculating the likelihood probability value corresponding to the attribute value sequence of the target attribute in each case under the condition that the user account is assumed to be each of the user account categories includes: Calculating candidate probability values ​​corresponding to the attribute value of each of the target attributes in each of the states under the condition that the user account is assumed to be of each of the user account categories; The likelihood probability value is calculated based on the candidate probability value.

6. The user identification method according to claim 1, characterized in that: Discretize the live broadcast service data to obtain the target data including: According to preset coding rules, the attribute values ​​of each attribute in the live broadcast service data are coded and integrated, wherein the states of the attribute values ​​of each attribute after coding and integration include: yes and no.

7. A user identification device, characterized in that: include: The preprocessing module is used to discretize the live broadcast service data to obtain the target data; An attribute determination module, used to determine a target attribute according to information gain values ​​of various attribute data in the target data, wherein the information gain value is used to characterize the influence of each attribute on determining a user account category, wherein the user account category includes: live streaming stealing users and non-live streaming stealing users; A probability calculation module, configured to obtain a target association relationship by calculating a posterior probability value of a user account being each of the user account categories, wherein the target association relationship is used to characterize the association relationship between the attribute value of the target attribute and the user account category, and obtaining the target association relationship by calculating a posterior probability value of a user account being each of the user account categories comprises: acquiring training sample data according to the target attribute, wherein the training sample data comprises a plurality of sample data of a plurality of user accounts, and each of the sample data comprises the target attribute and the user account category corresponding to the attribute value of the target attribute; based on the training sample data, training a classification model by calculating the posterior probability value of the user account being each of the user account categories under the condition that each attribute value in the attribute value sequence of the target attribute is in different states, thereby obtaining the target association relationship, wherein the attribute value sequence comprises all the target attributes, and the state of each attribute value of the target attribute comprises: yes and no; A category identification module is used to determine the user account category of each user account in the target data according to the target association relationship.

8. An electronic device, comprising a processor, characterized in that: The processor is used to run a program, wherein the program executes the user identification method according to any one of claims 1 to 6 when running.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the user identification method according to any one of claims 1 to 6 by running the computer program.

Images