A method, device, medium and computer equipment for identifying abnormal viewing users

Abstract

The present invention provides a method, device, medium and computer equipment for identifying abnormal viewing users, the method comprising: obtaining viewing index data of each live broadcast room; the viewing index comprising: viewing device information and viewing user ratio of each of the live broadcast rooms; constructing a corresponding first live broadcast viewing sequence based on the viewing index data; determining the suspicion value of each live broadcast room based on the first live broadcast viewing sequence; determining the abnormal live broadcast room based on the suspicion value, and determining the viewing suspicion jump radius of each user in the abnormal live broadcast room; identifying the abnormal viewing user based on the viewing suspicion jump radius; in this way, the suspicion value of the live broadcast room is used as an objective parameter to identify the abnormal viewing user, reduce subjectivity and improve recognition accuracy; and the present application takes into account that abnormal users often jump between abnormal live broadcast rooms, and therefore finally determines whether the user is an abnormal user based on the determined viewing suspicion jump radius of the user, which is close to the actual situation and further improves the recognition accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of live broadcast platforms, and in particular to a method, device, medium and computer equipment for identifying abnormal viewing users. Background Art

[0002] In live broadcast platforms, user viewing behavior is the most core behavior. In order to obtain a lot of resources or complete live broadcast tasks, anchors often have abnormal viewing behavior. This will damage the interests of other anchors, and bad money will drive out good money, thus causing the platform's live broadcast ecology to continue to deteriorate.

[0003] In the prior art, the mining method for identifying abnormal user viewing often adopts a rule-based method, formulating corresponding identification rules for users or live broadcast rooms. This method requires a lot of manual business experience and analysis, and is highly subjective, resulting in the inability to ensure identification accuracy. Summary of the invention

[0004] In response to the problems existing in the prior art, the embodiments of the present invention provide a method, apparatus, medium and computer equipment for identifying abnormal viewing users, which are used to solve the technical problem that in the prior art, when identifying abnormal viewing users, it is necessary to rely on a large amount of manual business experience, which is highly subjective and leads to the inability to ensure recognition accuracy.

[0005] In a first aspect, the present invention provides a method for identifying abnormal viewing users, the method comprising:

[0006] Obtain viewing indicator data for each live broadcast room; viewing indicators include: the number of viewing devices for each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IPs belong;

[0007] Constructing a corresponding first live viewing sequence based on the viewing indicator data;

[0008] Determine the suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence;

[0009] Determine an abnormal live broadcast room based on the suspicion value;

[0010] Determine the suspected viewing jump radius of each user in the abnormal live broadcast room;

[0011] Abnormal viewing users are identified based on the viewing suspicion jump radius.

[0012] In the above solution, the step of constructing a corresponding first live viewing sequence based on the viewing data includes:

[0013] For each viewing indicator, sorting the viewing indicator based on viewing time;

[0014] A sorted viewing indicator sequence is obtained, where the sorted viewing indicator sequence is a first live viewing sequence corresponding to the viewing indicator.

[0015] In the above solution, determining the suspicion value of each live broadcast room based on the first live broadcast viewing sequence includes:

[0016] According to the formula Determine the intermittent rate B of the first live viewing sequence ij ;

[0017] According to the formula Determine the repetition probability R of the first live viewing sequence ij ;

[0018] The suspicion value of the live broadcast room is determined based on the intermittent rate and repetition probability of the first live broadcast viewing sequence; wherein i is the sequence number of the live broadcast room, j is the sequence number of the viewing indicator in the live broadcast room, x is the total number of viewing indicators j in the first live broadcast viewing sequence, and σ ij is the standard deviation of the first live viewing sequence, and the m ij is the mean of the first live broadcast watching sequence; n is the length of the subsequence after the first live broadcast watching sequence is split into subsequences; t is the tth item in the first live broadcast watching sequence, is the value corresponding to the tth viewing indicator in the first live viewing sequence, is the average of the first n viewing indicators in the first live viewing sequence, is the average of the last n viewing indicators in the first live viewing sequence, is the standard deviation of the first n viewing indicators in the first live viewing sequence. is the standard deviation of the last n viewing indicators in the first live viewing sequence.

[0019] In the above solution, the determining the suspicion value of the live broadcast room based on the intermittent rate and repetition probability of the first live broadcast viewing sequence includes:

[0020] According to the formula Determine the suspicion value P of the live broadcast room i ; Wherein, i is the serial number of the live broadcast room, j is the serial number of the viewing indicator in the live broadcast room, and B ij is the intermittent rate of the first live viewing sequence, and the R ij is the repetition probability of the first live viewing sequence.

[0021] In the above solution, the determining of the user's viewing suspicion jump radius based on the suspicion value includes:

[0022] According to the formula Determine the user's suspected viewing jump radius R(u); where,

[0023] u is a user, s is the total number of sequence values ​​contained in the second live broadcast viewing sequence formed for each user, m is any sequence value in the second live broadcast viewing sequence, and r m is the live broadcast room number corresponding to the mth sequence value; is the suspicion value of the live broadcast room corresponding to the mth sequence value, Said is the proportion of the viewing time of the live broadcast room corresponding to the mth sequence value in the total viewing time of the user; k is an adjustment coefficient for adjusting the distribution of the viewing suspicion jump radius, and the value range of k is 1 to 5.

[0024] In the above solution, the step of identifying abnormal viewing users based on the viewing suspicion jump radius includes:

[0025] If it is determined that the suspected viewing jump radius is greater than a jump radius threshold, the user is determined to be an abnormal viewing user.

[0026] In a second aspect, the present invention provides a device for identifying abnormal viewing users, the device comprising:

[0027] An acquisition unit is used to acquire viewing index data of each live broadcast room; the viewing index includes: the number of viewing devices of each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IP belongs;

[0028] A construction unit, configured to construct a corresponding first live viewing sequence based on the viewing data;

[0029] A determination unit, configured to determine a suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence; determine an abnormal live broadcast room based on the suspicion value; and determine a suspected viewing jump radius of each user in the abnormal live broadcast room;

[0030] An identification unit is used to identify abnormal viewing users based on the viewing suspicion jump radius.

[0031] In the above scheme, the construction unit is specifically used for:

[0032] For each viewing indicator, sorting the viewing indicator based on viewing time;

[0033] A sorted viewing indicator sequence is obtained, wherein the sorted viewing indicator sequence and the live broadcast viewing sequence corresponding to the sorted viewing indicator sequence are the first live broadcast viewing sequence.

[0034] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the methods described in the first aspect.

[0035] In a fourth aspect, the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the methods described in the first aspect when executing the program.

[0036] The present invention provides a method, device, medium and computer equipment for identifying abnormal viewing users, the method comprising: obtaining viewing index data of each live broadcast room; the viewing index comprising: the number of viewing devices, the proportion of viewing users and the number of cities to which viewing IPs belong in each of the live broadcast rooms; constructing a corresponding first live viewing sequence based on the viewing index data; determining the suspicion value of each of the live broadcast rooms based on the first live viewing sequence; determining an abnormal live broadcast room based on the suspicion value; determining the viewing suspicion jump radius of each user in the abnormal live broadcast room; identifying abnormal viewing users based on the viewing suspicion jump radius; in this way, the suspicion value of the live broadcast room is first determined, and the suspicion value of the live broadcast room is used as an objective parameter to identify abnormal viewing users, thereby reducing subjectivity and improving recognition accuracy; and the present application takes into account that abnormal users often jump between abnormal live broadcast rooms, and therefore finally determines whether the user is an abnormal user based on the determined viewing suspicion jump radius of the user, which is close to the actual situation and has strong objectivity, and further improves recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0038] Figure 1 A schematic flow chart of a method for identifying abnormal viewing users provided by an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of the structure of a device for identifying abnormal viewing users provided by an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of the computer device structure provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to solve the technical problem that in the prior art, when identifying abnormal viewing users, a large amount of manual business experience is needed, which is highly subjective and leads to the inability to ensure the recognition accuracy, the present invention provides a method, device, medium and computer equipment for identifying abnormal viewing users.

[0043] In order to better understand the above technical scheme, the technical scheme of the embodiments of this specification is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of this specification and the specific features in the embodiments are detailed descriptions of the technical scheme of the embodiments of this specification, rather than limitations on the technical scheme of this specification. In the absence of conflict, the embodiments of this specification and the technical features in the embodiments can be combined with each other.

[0044] This embodiment provides a method for identifying abnormal viewing users. Figure 1 As shown, the method includes:

[0045] S110, obtaining viewing index data of each live broadcast room; the viewing index includes: viewing device information and viewing user ratio of each live broadcast room;

[0046] When users watch live broadcasts in a live broadcast room, corresponding behavior log data will be generated. The behavior log data records viewing indicator data. The viewing indicators include the number of viewing devices in each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IPs belong.

[0047] Some users have registered accounts on the live broadcast platform, and these users are collectively referred to as registered users; some users may not have registered accounts on the live broadcast platform, and these users are collectively referred to as visitors; the proportion of viewing users is the quotient between the number of registered users entering each live broadcast room and the total number.

[0048] In order to find abnormal viewing users, it is necessary to collect statistics on viewing related indicators. The reason for selecting the above viewing indicators in this embodiment is that when watching in the live broadcast room, the number of viewing devices will inevitably increase; in order to reduce account costs, more visitor devices will be used, so the proportion of visitors will increase; at the same time, the IP used will concentrate the cities where the viewing IPs are viewed, so the number of cities to which the viewing IPs belong will actually decrease.

[0049] This embodiment can collect the viewing index data of each live broadcast room within a preset time period and with a preset collection period. The collection period can be set according to actual conditions, for example, 1 minute, 5 minutes or 60 minutes, and there is no limitation here.

[0050] This step selects the above viewing indicators with full consideration of the actual situation, ensuring the objectivity and authenticity of the basic data, and further ensuring the objectivity of subsequent identification.

[0051] S111, constructing a corresponding first live broadcast viewing sequence based on the viewing indicator data;

[0052] After the viewing index data of each live broadcast room is obtained, for each live broadcast room, a corresponding first live broadcast viewing sequence can be constructed based on the viewing index data.

[0053] In this embodiment, constructing a corresponding first live viewing sequence based on viewing indicator data includes:

[0054] For each viewing indicator, the viewing indicators are sorted based on the viewing time;

[0055] A sorted viewing indicator sequence is obtained, where the sorted viewing indicator sequence is a first live viewing sequence corresponding to the viewing indicator.

[0056] It can be seen that each viewing indicator has a corresponding first viewing live broadcast sequence.

[0057] For example, for a certain live broadcast room, when constructing the first live broadcast viewing sequence corresponding to the number of viewing devices, if the number of devices collected every 5 minutes within half an hour is 20, 30, 40, 35, 20, 30 respectively: then the corresponding first live broadcast viewing sequence is {20, 30, 40, 35, 20, 30}.

[0058] This step constructs a corresponding first live viewing sequence for each live broadcast room. The first live viewing sequence is the basis for risk identification, and anomalies can be identified based on the sequence.

[0059] S112, determining a suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence;

[0060] After all first live viewing sequences of each live broadcast room are determined, the suspicion value of each live broadcast room is determined based on the first live viewing sequence.

[0061] In this embodiment, determining the suspicion value of each live broadcast room based on the first live broadcast viewing sequence includes:

[0062] According to the formula Determine the intermittent rate B of the first live viewing sequence ij ;

[0063] According to the formula Determine the repetition probability R of the first live viewing sequence ij ;

[0064] Determining the suspicion value of the live broadcast room based on the intermittent rate and the repetition probability of the first live broadcast viewing sequence;

[0065] Among them, i is the serial number of the live broadcast room, j is the serial number of the viewing index in the live broadcast room, and each viewing index j corresponds to a first viewing live broadcast sequence, σ ij is the standard deviation of the first live broadcast sequence, x is the total number of sequence values ​​contained in the first live broadcast sequence, m ij is the mean of the first live broadcast watching sequence; n is the length of the subsequence after the first live broadcast watching sequence is split into subsequences, n is a constant, and its value is related to the length of the first live broadcast watching sequence, and n is generally 1 / 2 of the length of the first live broadcast watching sequence; t is the tth item in the first live broadcast watching sequence, is the value corresponding to the tth viewing indicator in the first live viewing sequence, is the value corresponding to the x-n+tth viewing indicator in the first live viewing sequence; is the average of the first n viewing indicators in the first live viewing sequence, is the average of the last n viewing indicators in the first live viewing sequence, is the standard deviation of the first n viewing indicators in the first live viewing sequence, is the standard deviation of the last n viewing indicators in the first live viewing sequence.

[0066] Above The principle of the formula is: the standard deviation of the sequence represents the volatility of the corresponding indicator of the sequence, while the intermittency represents the uniformity of the overall distribution of the sequence. If the overall distribution of the sequence is uniform, then the relative interval between the standard deviation + mean and the standard deviation - mean will be relatively small, otherwise it will be relatively large. Therefore, the relative ratio of the standard deviation - mean and the standard deviation + mean is used to measure the intermittent rate.

[0067] Above The principle of the formula is: for each viewing indicator, the corresponding first viewing live sequence is split into a combination of the first n viewing indicator subsequences and the last n viewing indicator subsequences. The stronger the correlation between the two subsequences, the stronger the repeatability of the sequence. Therefore, the formula Represents the values ​​of the first n viewing indicators in the first subsequence and the corresponding values ​​of the last n viewing indicators in the second subsequence The correlation between the two subsequences is measured by calculating the sum and average of the n viewing indicators.

[0068] For example, if a first live broadcast sequence is {20, 30, 40, 35, 20, 30}, which contains 6 sequence values, the standard deviation σ of the first live broadcast sequence can be calculated. ij is 8.37, mean m ij It is 30.8, x=6.

[0069] Take n = 3,

[0070] then:

[0071] The average of the top 3 viewing indicators is 30, standard deviation It is 8.16;

[0072] The average of the last three viewing indicators is 31.7, standard deviation It is 8.5;

[0073] It can be calculated that:

[0074]

[0075]

[0076] For a live broadcast room with the behavior of swiping views, since a certain number of task orders need to be completed quickly, intensive behavior will occur in a short period of time, and then there will be a longer idle period. Therefore, the higher the intermittent rate, the more abnormal the live broadcast room is; and since swiping volume is done by scripts, there will inevitably be periodic fluctuations. Therefore, the greater the repeatability, the more abnormal the live broadcast room is.

[0077] Then in this embodiment, the suspicion value of the live broadcast room is determined based on the intermittent rate and repetition probability of the first live broadcast viewing sequence, including:

[0078] According to the formula Determine the suspicion value P of the live broadcast room i ; where i is the serial number of the live broadcast room, j is the serial number of the viewing indicator in the live broadcast room, B ij is the intermittent rate of the first live viewing sequence, R ij is the repetition probability of the first viewing live sequence.

[0079] The principle of the above formula is: the product of the intermittency and repeatability of the sequence represented by the viewing index j represents the suspicion value of the sequence.

[0080] For example: Assume there are two viewing indicators, and the B calculated when j = 1 i1 =0.5, R i1 =0.3; B calculated when j=2 i2 =0.3, R i2 =0.2

[0081] So: P i =max{0.5*0.3,0.3*0.2}=0.15.

[0082] This step determines the suspicion value of the live broadcast room through the intermittent rate and repetition probability of the live broadcast sequence, and determines the suspicion value of each live broadcast room, laying the foundation for the subsequent identification of abnormal users.

[0083] S113, determining an abnormal live broadcast room based on the suspicion value, and determining a suspected viewing jump radius of each user in the abnormal live broadcast room;

[0084] After the suspicion value of the live broadcast room is determined, the abnormal live broadcast room with brushing viewing behavior can be determined according to the suspicion value. For example, if the suspicion value of the live broadcast room is 0.15, which is greater than the suspicion threshold of 0.1, the live broadcast room can be determined as an abnormal live broadcast room.

[0085] However, users in abnormal live broadcast rooms cannot be considered all abnormal users, and there may be some normal users. For users who watch a lot of videos, they often use one account to complete multiple order-brushing tasks, so such users will jump between abnormal live broadcast rooms. For normal users, although they may watch some abnormal live broadcast rooms, most of the time they watch normal live broadcast rooms, so they will jump between abnormal and normal live broadcast rooms. Based on this, this step also needs to be based on the suspected jump radius of each user in the abnormal live broadcast room.

[0086] Specifically, since each user should also have a corresponding second live broadcast viewing sequence, then the second live broadcast viewing sequence corresponding to each user is determined according to the formula Determine the user's suspected viewing jump radius R(u); where,

[0087] u is a user, s is the total number of sequence values ​​contained in the second live broadcast viewing sequence formed for each user, m is any sequence value in the second live broadcast viewing sequence, r m is the live broadcast room number corresponding to the mth sequence value; is the suspicion value of the live broadcast room corresponding to the mth sequence value, is the proportion of the viewing time of the live broadcast room corresponding to the mth sequence value in the total viewing time of the user; k is the adjustment coefficient used to adjust the distribution of the viewing suspicion jump radius, and the value range of k is 1 to 5.

[0088] It is worth noting that the adjustment coefficient k is used to adjust the strictness of risk identification (which can also be understood as the degree of leniency or strictness). The larger the k, the larger the suspect jump radius, the more users are identified, and the higher the identification coverage; the smaller the k, the smaller the suspect jump radius, the fewer users are identified, and the higher the identification accuracy. The value range of k is determined according to the requirements of the accuracy and coverage of the specific business scenario, and can be set based on the actual situation, and is not limited here.

[0089] The principle of the above formula: The above formula consists of two parts. If a user frequently jumps from one high-risk live broadcast room to another high-risk live broadcast room, it means that the user's abnormal viewing risk is very high, so the The risk difference of the jump is represented by ; the relationship between this risk difference and the final risk value is not linear, so an exponential function is used here to mathematically transform the risk difference. The viewing time ratio is used as the weight for the summation, and the risk difference of each jump is weighted summed and averaged to obtain the first term of the formula This item represents the average jump risk. The larger the value of this item, the greater the overall suspicion score. The other item in the formula is the weighted average suspicion score of the live broadcast room. The more suspicious the live broadcast room the user watches, the higher the user's suspicion score will be.

[0090] For example:

[0091] Assume that the live broadcast room numbers corresponding to the sequence values ​​(3 sequence values) in the second live broadcast viewing sequence of a user u are sorted by time, namely A, B, and C: the live broadcast room number corresponding to the first sequence value is A, the live broadcast room number corresponding to the second sequence value is B, and the live broadcast room number corresponding to the third sequence value is C; among them,

[0092] The viewing time of live broadcast room A accounts for 0.3, and the suspicion value of live broadcast room A is 0.2; the viewing time of live broadcast room B accounts for 0.3, and the suspicion value of live broadcast room B is 0.4; the viewing time of live broadcast room C accounts for 0.4, and the suspicion value of live broadcast room C is 0.1. Let k = 1, then:

[0093]

[0094] S114: Identify abnormal viewing users based on the suspected viewing jump radius.

[0095] After the suspected viewing jump radius is determined, abnormal viewing users are identified based on the suspected viewing jump radius.

[0096] In this embodiment, identifying abnormal viewing users based on the viewing suspicion jump radius includes:

[0097] If it is determined that the suspected viewing jump radius is greater than the jump radius threshold, the user is determined to be an abnormal viewing user.

[0098] Continuing with the above example, if the jump radius threshold is 0.015, the viewing suspicion jump radius calculated above is 0.021, and the user u will be identified as an abnormal viewing user.

[0099] It is worth noting that: if you need to prioritize the accuracy of recognition and avoid false positives as much as possible, then increase the jump radius threshold based on the actual scenario; if you need to cover as many abnormal users as possible, you can lower the jump radius threshold based on the actual situation.

[0100] The method for identifying abnormal viewing users provided by the present embodiment first determines the suspicion value of the live broadcast room, and uses the suspicion value of the live broadcast room as an objective parameter to identify abnormal viewing users, thereby reducing subjectivity and improving recognition accuracy. In addition, the present application takes into account that abnormal users often jump between abnormal live broadcast rooms, and therefore ultimately determines whether the user is an abnormal user based on the user's viewing suspicion jump radius, which is close to the actual situation and has strong objectivity, thereby further improving recognition accuracy.

[0101] Based on the same inventive concept, this embodiment also provides a device for identifying abnormal viewing users, such as Figure 2 As shown, the device comprises:

[0102] The acquisition unit 21 is used to acquire the viewing index data of each live broadcast room; the viewing index includes: the number of viewing devices, the proportion of viewing users and the number of cities to which the viewing IPs belong in each of the live broadcast rooms;

[0103] A construction unit 22, configured to construct a corresponding first live viewing sequence based on the viewing data;

[0104] A determination unit 23 is used to determine an abnormal live broadcast room based on the suspicion value; and determine a suspected viewing jump radius of each user in the abnormal live broadcast room;

[0105] The identification unit 24 is used to identify abnormal viewing users based on the suspected viewing jump radius.

[0106] The specific functions of the above units can be found in the corresponding descriptions in the above method embodiments, which will not be described in detail here. Since the device introduced in the embodiment of the present invention is a device used to implement the method of the embodiment of the present invention, based on the method introduced in the embodiment of the present invention, those skilled in the art can understand the specific structure and deformation of the device, so they will not be described in detail here. All devices used in the method of the embodiment of the present invention belong to the scope of protection of the present invention.

[0107] Based on the same inventive concept, this embodiment provides a computer device 300, such as Figure 3 As shown, it includes a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented:

[0108] Obtain viewing indicator data for each live broadcast room; viewing indicators include: the number of viewing devices for each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IPs belong;

[0109] Constructing a corresponding first live viewing sequence based on the viewing indicator data;

[0110] Determine the suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence;

[0111] Determine an abnormal live broadcast room based on the suspicion value; determine a suspected viewing jump radius of each user in the abnormal live broadcast room;

[0112] Abnormal viewing users are identified based on the viewing suspicion jump radius.

[0113] In a specific implementation process, when the processor 320 executes the computer program 311, any implementation method in the foregoing embodiments can be implemented.

[0114] Since the computer device introduced in this embodiment is a device used to implement a method for identifying abnormal viewing users in an embodiment of the present application, based on the method introduced in the aforementioned embodiment of the present application, a person skilled in the art can understand the specific implementation of the computer device of the present embodiment and its various variations, so how the server implements the method in the embodiment of the present application is not described in detail here. As long as the device used by a person skilled in the art to implement the method in the embodiment of the present application is within the scope of protection of the present application.

[0115] Based on the same inventive concept, this embodiment provides a computer-readable storage medium 400, such as Figure 4 As shown, a computer program 411 is stored thereon, and when the computer program 411 is executed by a processor, the following steps are implemented:

[0116] Obtain viewing indicator data for each live broadcast room; viewing indicators include: the number of viewing devices for each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IPs belong;

[0117] Constructing a corresponding first live viewing sequence based on the viewing indicator data;

[0118] Determine the suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence;

[0119] Determine an abnormal live broadcast room based on the suspicion value; determine a suspected viewing jump radius of each user in the abnormal live broadcast room;

[0120] Abnormal viewing users are identified based on the viewing suspicion jump radius.

[0121] In the specific implementation process, when the computer program 411 is executed by a processor, any implementation method in the aforementioned embodiments can be implemented.

[0122] The method, device, medium and computer equipment provided by the present invention for identifying abnormal viewing users can bring at least the following beneficial effects:

[0123] The present invention provides a method, device, medium and computer equipment for identifying abnormal viewing users, the method comprising: obtaining viewing index data of each live broadcast room; the viewing index comprising: the number of viewing devices, the proportion of viewing users and the number of cities to which viewing IPs belong in each of the live broadcast rooms; constructing a corresponding first live viewing sequence based on the viewing index data; determining the suspicion value of each of the live broadcast rooms based on the first live viewing sequence; determining an abnormal live broadcast room based on the suspicion value; determining the viewing suspicion jump radius of each user in the abnormal live broadcast room; identifying abnormal viewing users based on the viewing suspicion jump radius; in this way, the suspicion value of the live broadcast room is first determined, and the suspicion value of the live broadcast room is used as an objective parameter to identify abnormal viewing users, thereby reducing subjectivity and improving recognition accuracy; and the present application takes into account that abnormal users often jump between abnormal live broadcast rooms, and therefore finally determines whether the user is an abnormal user based on the determined viewing suspicion jump radius of the user, which is close to the actual situation and has strong objectivity, and further improves recognition accuracy.

[0124] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0125] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0128] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0129] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying abnormal viewing users, characterized in that: The method comprises: Obtain viewing indicator data for each live broadcast room; viewing indicators include: the number of viewing devices for each live broadcast room, the proportion of viewing users, and the number of cities to which the viewing IPs belong; Constructing a corresponding first live viewing sequence based on the viewing indicator data; Determine the suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence; Determine an abnormal live broadcast room based on the suspicion value; Determine the suspected viewing jump radius of each user in the abnormal live broadcast room; Identifying abnormal viewing users based on the viewing suspicion jump radius; The constructing a corresponding first live viewing sequence based on the viewing data includes: For each viewing indicator, sorting the viewing indicator based on viewing time; Obtaining a sorted viewing indicator sequence, wherein the sorted viewing indicator sequence is a first live viewing sequence corresponding to the viewing indicator; The determining the suspicion value of each of the live broadcast rooms based on the first live broadcast viewing sequence includes: According to the formula Determine the intermittent rate B of the first live viewing sequence ij ; According to the formula Determine the repetition probability R of the first live viewing sequence ij ; The suspicion value of the live broadcast room is determined based on the intermittent rate and repetition probability of the first live broadcast viewing sequence; wherein i is the sequence number of the live broadcast room, j is the sequence number of the viewing indicator in the live broadcast room, x is the total number of viewing indicators j in the first live broadcast viewing sequence, and σ ij is the standard deviation of the first live viewing sequence, and the m ij is the mean of the first live broadcast watching sequence; n is the length of the subsequence after the first live broadcast watching sequence is split into subsequences; t is the tth item in the first live broadcast watching sequence, is the value corresponding to the tth viewing indicator in the first live viewing sequence, is the average of the first n viewing indicators in the first live viewing sequence, is the average of the last n viewing indicators in the first live viewing sequence, is the standard deviation of the first n viewing indicators in the first live viewing sequence. is the standard deviation of the last n viewing indicators in the first live viewing sequence; The determining the user's viewing suspicion jump radius based on the suspicion value includes: According to the formula Determine the user's suspected viewing jump radius R(u); where, u is a user, s is the total number of sequence values ​​contained in the second live broadcast viewing sequence formed for each user, m is any sequence value in the second live broadcast viewing sequence, and r m is the live broadcast room number corresponding to the mth sequence value; is the suspicion value of the live broadcast room corresponding to the mth sequence value, Said is the proportion of the viewing time of the live broadcast room corresponding to the mth sequence value in the total viewing time of the user; k is the adjustment coefficient used to adjust the distribution of the viewing suspicion jump radius.

2. The method according to claim 1, characterized in that The determining the suspicion value of the live broadcast room based on the intermittent rate and repetition probability of the first live broadcast viewing sequence includes: According to the formula Determine the suspicion value P of the live broadcast room i ; Wherein, i is the serial number of the live broadcast room, j is the serial number of the viewing indicator in the live broadcast room, and B ij is the intermittent rate of the first live viewing sequence, and the R ij is the repetition probability of the first live viewing sequence.

3. The method according to claim 1, characterized in that The identifying of abnormal viewing users based on the viewing suspicion jump radius includes: If it is determined that the suspected viewing jump radius is greater than a jump radius threshold, the user is determined to be an abnormal viewing user.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

5. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 3 is implemented.

Images