A method, device, medium and equipment for identifying a live broadcaster with ticket-scalping behavior

Abstract

The present invention provides a method, device, medium and equipment for identifying an anchor with ticket stuffing behavior. The method includes: determining a voting account feature tree according to the voting account feature information of a target anchor; obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; determining the ticket stuffing suspicion score corresponding to the target non-leaf node based on the voting account feature subset; identifying the target anchor based on the ticket stuffing suspicion score. In this way, the voting account feature information is first selected as the basic identification data. Since the above information is closely related to the actual ticket stuffing behavior, the accuracy of the basic identification data can be ensured. All possible voting account feature subsets are included in the voting account feature tree. When determining the ticket stuffing suspicion score according to the voting account feature subset, comprehensiveness and objectivity can be ensured, the accuracy of the ticket stuffing suspicion score can be ensured, and further the identification accuracy of the anchor with ticket stuffing behavior can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of risk control of live broadcast platforms, and particularly relates to a method, device, medium and equipment for identifying a live broadcast host with a ticket-scalping behavior. Background Art

[0002] In a live broadcast platform, there are some host activities where users can vote for the host, and the host with a higher vote count will receive corresponding rewards. For the sake of activity rewards, some hosts often use improper means to scalp tickets.

[0003] Therefore, in order to ensure the interests of the platform, it is necessary to identify the hosts with ticket-scalping behavior in the host activities. In the prior art, generally, the vote count of the host in the previous voting activity or the vote count of hosts of the same level is compared. If there is a significant increase, it is considered that the host has a suspicion of ticket-scalping. Although this method is easy to implement, the identification error is particularly large and the accuracy cannot be ensured. Summary of the Invention

[0004] In view of the problems existing in the prior art, an embodiment of the present invention provides a method, device, medium and equipment for identifying a live broadcast host with a ticket-scalping behavior, which is used to solve the technical problem that the host with a ticket-scalping behavior cannot be accurately identified in the prior art.

[0005] In a first aspect, the present invention provides a method for identifying a live broadcast host with a ticket-scalping behavior, the method comprising:

[0006] Determining a voting account feature tree according to the voting account feature information of a target host; the voting account feature information includes: device attribute information, device IP information and voting user attribute information;

[0007] Obtaining each non-leaf node of the voting account feature tree and a voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information or voting user attribute information;

[0008] For a target non-leaf node, determining a ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset; the target non-leaf node is any one of the non-leaf nodes;

[0009] Identifying the target host based on the ticket-scalping suspicion score.

[0010] Optionally, the determining a voting account feature tree according to the voting account feature information of a target host includes:

[0011] Generating a root node of the voting account feature tree;

[0012] Obtain the first feature information included in the device attribute information, the first feature value corresponding to the first feature information, the second feature information included in the device IP information, the second feature value corresponding to the second feature information, the third feature information included in the voting user attribute information, and the third feature value corresponding to the third feature information;

[0013] Traverse the first feature value, the second feature value, and the third feature value, form feature value sets at different levels according to the first feature value, the second feature value, and the third feature value, and determine the corresponding layer nodes of the voting account feature tree according to the feature value sets at different levels; where

[0014] The number of feature values included in the current layer node of the voting account feature tree is greater than the number of feature values included in the node of the previous layer of the current layer node.

[0015] Optionally, obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node includes:

[0016] Taking the first layer node of the voting account feature tree as a non-leaf node;

[0017] Taking the nodes other than the first layer node as the voting account feature subset of each non-leaf node.

[0018] Optionally, determining the ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset includes:

[0019] According to the formula Determine the ticket-scalping suspicion score ps(x) corresponding to the target non-leaf node; where

[0020] The x is the target non-leaf node, the y i is the voting account feature subset, the i is the serial number of the voting account feature subset, the n is the total number of the voting account feature subsets, the v(y i ) is the actual number of the voting account feature subset, the f(y i ) is the predicted number of the voting account feature subset, and the a(y i ) is the inferred number of the voting account feature subset.

[0021] Optionally, when the voting account feature subset is the subset corresponding to the target non-leaf node, the method further includes:

[0022] According to the formula Determine the inferred number a(y i ) of the voting account feature subset; where

[0023] where f(y i ) is the predicted quantity of the subset of voting account features, y i is the subset of voting account features, i is the serial number of the subset of voting account features, f(x) is the predicted quantity of the non-leaf node, v(x) is the actual quantity of the non-leaf node, and x is the target non-leaf node.

[0024] Optionally, if the subset of voting account features does not belong to the subset of the target non-leaf node, the method further includes:

[0025] Determining the inferred quantity a(y i ) of the subset of voting account features according to the formula a(y i ) = f(y i ); where f(y i ) is the predicted quantity of the subset of voting account features, y i is the subset of voting account features, and i is the serial number of the subset of voting account features.

[0026] Optionally, the identifying the target anchor based on the suspected vote rigging score includes:

[0027] Obtaining the suspected vote rigging scores of all target non-leaf nodes;

[0028] If it is determined that there is any suspected vote rigging score greater than the detection threshold, determining that the target anchor is an anchor with vote rigging behavior.

[0029] In a second aspect, the present invention provides an apparatus for identifying an anchor with vote rigging behavior, the apparatus including:

[0030] A first determination unit, configured to determine a voting account feature tree according to the voting account feature information of a target anchor; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information;

[0031] An obtaining unit, configured to obtain each non-leaf node of the voting account feature tree and the subset of voting account features of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information;

[0032] A second determination unit, configured to determine, for a target non-leaf node, a suspected vote rigging score corresponding to the target non-leaf node based on the subset of voting account features; the target non-leaf node is any one of the non-leaf nodes;

[0033] An identification unit, configured to identify the target anchor based on the suspected vote rigging score.

[0034] In a third aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method described in any one of the first aspects is implemented.

[0035] In a fourth aspect, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the method described in any one of the first aspects is implemented.

[0036] The present invention provides a method, device, medium, and device for identifying an anchor with a ticket-scalping behavior. The method includes: determining a voting account feature tree according to the voting account feature information of a target anchor; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information; obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information; for a target non-leaf node, based on the voting account feature subset, determining a ticket-scalping suspicion score corresponding to the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes; identifying the target anchor based on the ticket-scalping suspicion score; thus, first, the voting account feature information is selected as the basic identification data. Since the above information is closely related to the actual ticket-scalping behavior, the accuracy of the basic identification data can be ensured; all possible voting account feature subsets are included in the voting account feature tree. When determining the ticket-scalping suspicion score according to the voting account feature subset, comprehensiveness and objectivity can be ensured, the accuracy of the ticket-scalping suspicion score can be ensured, and further the identification accuracy of an anchor with a ticket-scalping behavior can be improved. Description of the Drawings

[0037] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0038] Figure 1 It is a schematic flowchart of the method for identifying an anchor with a ticket-scalping behavior provided by an embodiment of the present invention;

[0039] Figure 2 It is a schematic structural diagram of the voting account feature tree provided by an embodiment of the present invention;

[0040] Figure 3 It is a schematic structural diagram of the device for identifying an anchor with a ticket-scalping behavior provided by an embodiment of the present invention;

[0041] Figure 4Schematic diagram of the computer device provided by the embodiment of the present invention;

[0042] Figure 5 Schematic diagram of the computer-readable storage medium structure provided by the embodiment of the present invention. Detailed implementation manners

[0043] In order to solve the technical problem in the prior art that the host with the ticket-scalping behavior cannot be accurately identified, the present invention provides a method, device, medium and equipment for identifying the host with the ticket-scalping behavior.

[0044] In order to better understand the above technical solution, the technical solution of the embodiments of this specification will be described in detail below through the drawings and specific embodiments. It should be understood that the specific features in the embodiments of this specification and the embodiments are detailed descriptions of the technical solution of the embodiments of this specification, rather than limitations on the technical solution of this specification. Without conflict, the technical features in the embodiments of this specification and the embodiments can be combined with each other.

[0045] This embodiment provides a method for identifying a host with a ticket-scalping behavior, as Figure 1 shown, the method includes:

[0046] S110, determining a voting account feature tree according to the voting account feature information of the target host; the voting account feature information includes: device attribute information, device IP information and voting user attribute information;

[0047] When ticket-scalping is performed on the live broadcast platform, the inventors of the present application found that the voting account feature information can best reflect the abnormality. Therefore, for any target host, the voting account information under the target host can be obtained, and the voting account feature information includes: device attribute information, device IP information and voting user attribute information.

[0048] Among them, the device attribute information includes first feature information. For example, the first feature information may include: device model, device brand, etc.; each first feature information includes at least one first feature value. For example, the first feature value of the device model may be tablet, smart phone, etc.; the first feature value of the device brand may include: brand 1, brand 2, etc.

[0049] The device IP information includes second feature information. For example, the second feature information may include: IP province or IP city, etc.; each second feature information includes at least one second feature value. For example, the second feature value of the IP province may include: Hubei, Guangdong, etc.; the second feature value of the IP city may include: Wuhan, Shanghai, Beijing, etc.

[0050] The voting user attribute information includes third feature information, and the third feature information is: user level, whether being a fan of the target anchor, etc.; each third feature information includes at least one third feature value. For example, the third feature values corresponding to the user level may include: level one, level two, etc.

[0051] After the feature values of each voting account feature information are determined, the voting account feature tree can be determined according to the voting account feature information.

[0052] In an alternative embodiment, determining the voting account feature tree according to the voting account feature information for the target anchor includes:

[0053] Generate the root node of the voting account feature tree;

[0054] Obtain the first feature information included in the device attribute information, the first feature value corresponding to the first feature information, the second feature information included in the device IP information, the second feature value corresponding to the second feature information, the third feature information included in the voting user attribute information, and the third feature value corresponding to the third feature information;

[0055] Traverse the first feature value, the second feature value, and the third feature value, form different hierarchical feature value sets according to the first feature value, the second feature value, and the third feature value, and determine the corresponding layer nodes of the voting account feature tree according to the different hierarchical feature value sets; where

[0056] The number of feature values included in the current layer node of the voting account feature tree is greater than the number of feature values included in the node of the previous layer of the current layer node.

[0057] For example, in order to better understand the structure of the voting account feature tree, it is assumed here that there are two feature values: the first feature value and the second feature value; where the first feature value is brand 1, brand 2; the second feature value is Wuhan, Shanghai, Beijing.

[0058] Then it can be referred to Figure 2 , the root node is root, and the first layer nodes are: brand 1, brand 2, Wuhan, Shanghai, Beijing; then the second layer nodes can be (brand 1, Wuhan), (brand 1, Shanghai), (brand 1, Beijing), (brand 2, Wuhan), (brand 2, Shanghai), (brand 2, Beijing).

[0059] Among them, the root node root is the parent node of the first layer nodes, the first layer nodes are non-leaf nodes, and the second layer nodes are the voting account feature subsets of the first layer nodes. The number of feature values in the second layer nodes is greater than the number of feature values included in the first layer nodes.

[0060] It should be noted that each first-layer node has a corresponding subset of voting accounts. For example, the subsets of voting account features for Brand 1 are (Brand 1, Wuhan), (Brand 1, Shanghai), and (Brand 1, Beijing), and the subset of voting accounts in Wuhan is (Brand 1, Wuhan) and (Brand 2, Wuhan).

[0061] In this step, a voting account feature tree is determined. The voting account feature tree contains all possible subsets of voting account features. Subsequently, when determining the ticket-scamming suspicion score based on the subsets of voting account features, comprehensiveness and accuracy can be ensured.

[0062] S111, obtain each non-leaf node of the voting account feature tree and the subset of voting account features for each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information.

[0063] After the voting account feature tree is determined, obtain each non-leaf node of the voting account feature tree and the subset of voting account features for each non-leaf node; the non-leaf nodes include one of device attribute information, device IP information, or voting user attribute information. Specifically, the non-leaf nodes are the characteristic values included in the device attribute information, device IP information, or voting user attribute information.

[0064] As described above, the voting account feature tree contains multiple layers of nodes. In an optional embodiment, obtaining each non-leaf node of the voting account feature tree and the subset of voting account features for each non-leaf node includes:

[0065] Take the first-layer nodes of the voting account feature tree as non-leaf nodes;

[0066] Take the nodes other than the first-layer nodes as the subsets of voting account features for each non-leaf node.

[0067] That is, when the first-layer nodes are: Brand 1, Brand 2, Wuhan, Shanghai, and Beijing; and the second-layer nodes can be (Brand 1, Wuhan), (Brand 1, Shanghai), (Brand 1, Beijing), (Brand 2, Wuhan), (Brand 2, Shanghai), and (Brand 2, Beijing), the non-leaf nodes are Brand 1, Brand 2, Wuhan, Shanghai, and Beijing; the subsets of voting account features for the non-leaf nodes include: (Brand 1, Wuhan), (Brand 1, Shanghai), (Brand 1, Beijing), (Brand 2, Wuhan), (Brand 2, Shanghai), and (Brand 2, Beijing).

[0068] In this step, by determining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node, the actual value and predicted value of the voting account feature subset participating in the calculation of the ticket-scraping suspicion score can be determined, and then the ticket-scraping suspicion score of the non-leaf node can be determined according to the actual value and predicted value of the voting account feature subset.

[0069] S112. For the target non-leaf node, based on the voting account feature subset, determine the ticket-scraping suspicion score corresponding to the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes;

[0070] For the target non-leaf node, based on the voting account feature subset, determine the ticket-scraping suspicion score corresponding to the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes;

[0071] In an optional embodiment, determining the ticket-scraping suspicion score corresponding to the target non-leaf node based on the voting account feature subset includes:

[0072] According to the formula Determine the ticket-scraping suspicion score ps(x) corresponding to the target non-leaf node; where

[0073] x is the target non-leaf node, y i is the voting account feature subset, i is the serial number of the voting account feature subset, n is the total number of voting account feature subsets, v(y i ) is the actual number of the voting account feature subset, f(y i ) is the predicted number of the voting account feature subset, a(y i ) is the inferred number of the voting account feature subset.

[0074] Among them, the determination method of the predicted number can be as follows:

[0075] It is possible to determine each reference anchor with the same level as the target anchor, obtain the historical account number of each reference anchor that meets the voting account feature subset in the historical voting activity, and use the historical account number as the predicted number of the voting account feature subset of the target anchor.

[0076] The actual number is the actual account number that participates in the voting activity of the target anchor and meets the voting account feature subset.

[0077] The inferred number can be understood as the account number that meets the voting account feature subset inferred according to the parent node (non-leaf node) of the voting account feature subset.

[0078] The principle of the above formula is: calculate the actual number v(y i ) and the inferred number a(y i) The difference between them is measured by taking the square root of the sum of the squares of the differences between the two; similarly, calculate the actual quantity v(y ), and the difference between the predicted quantity f(y i ) and the predicted quantity f(y i ). The difference between the actual quantity and the inferred quantity represents the difference when there is no anomaly and represents normal fluctuations; while the difference between the predicted quantity and the actual quantity represents the difference after an actual anomaly occurs. Therefore, divide the above results. If the anomaly is greater, then this ratio will be smaller, and the suspicion score of ticket stuffing will be greater.

[0079] Among them, in an optional embodiment, if the subset of voting account features is the subset corresponding to the target non-leaf node, the method further includes:

[0080] Determine the inferred quantity a(y ) of the subset of voting account features according to the formula i ; where

[0081] f(y i ) is the predicted quantity of the subset of voting account features, y i is the subset of voting account features, i is the serial number of the subset of voting account features, f(x) is the predicted quantity of the non-leaf node, v(x) is the actual quantity of the non-leaf node, and x is the target non-leaf node.

[0082] The principle of the above formula is: for the subset of voting account features belonging to the target non-leaf node x, the overall change of x has a spreading effect on the subset. Therefore, it is necessary to adjust the predicted quantity of x to obtain the predicted quantity after removing the anomaly. The adjustment principle is: each y i has the same unit contribution to the anomaly. Therefore, the abnormal change quantity of y i can be obtained by multiplying the abnormal change quantity f(x) - v(x) of x by the proportion of the predicted quantity of y . Subtract this abnormal change quantity from the original predicted quantity f(y i ) to obtain the corrected inferred quantity a(y i ). i )

[0083] If the subset of voting account features does not belong to the subset of the target non-leaf node, the method further includes:

[0084] Determine the inferred quantity a(y i ) of the subset of voting account features according to the formula a(y i ) = f(y i ); where f(y i ) is the predicted quantity of the subset of voting account features, y iIt is a subset of voting account features, and i is the serial number of the subset of voting account features.

[0085] The principle of this formula is as follows: For the subset of voting account features that do not belong to the target non-leaf node x, the overall change of x has no impact on the subset. Therefore, it is inferred that the quantity a(y i ) can be directly equivalent to the predicted quantity a(y i ).

[0086] For example, continuing with the example above, assume there are two feature values: the first feature value and the second feature value; among them, the first feature value is Brand 1, Brand 2; the second feature value is Wuhan, Shanghai, Beijing; then x can include: x1 is Brand 1, x2 is Brand 2, x3 is Wuhan, x4 is Beijing, x5 is Shanghai; y i can include: y1 = (Brand 1, Wuhan), y2 = (Brand 1, Shanghai), y3 (Brand 1, Beijing), y4 = (Brand 2, Wuhan), y5 = (Brand 2, Shanghai), y6 = (Brand 2, Beijing). The predicted quantities of each x and each y i are shown in Table 1:

[0087] Table 1

[0088] IP City = Wuhan IP City = Beijing IP City = Shanghai Predicted Brand Quantity Brand 1 20 15 10 45 Brand 2 10 25 20 55 Predicted IP Quantity 30 40 30

[0089] As can be seen from Table 1, f(y1) is 20, f(y2) is 15, f(y3) is 15, f(y4) is 10, f(y5) is 25, f(y6) is 20; f(x1) is 45, f(x2) is 55, f(x3) is 30, f(x4) is 40, f(x5) is 30.

[0090] The actual quantities of each x and each y i are shown in Table 2:

[0091] Table 2

[0092] IP City = Wuhan IP City = Beijing IP City = Shanghai Actual Brand Quantity Brand 1 14 9 10 33 Brand 2 7 15 20 42 Actual IP Quantity 21 24 30

[0093] As can be seen from Table 2, v(y1) is 14, v(y2) is 9, v(y3) is 10, v(y4) is 7, v(y5) is 15, v(y6) is 20; v(x1) is 33, v(x2) is 42, v(x3) is 21, v(x4) is 24, v(x5) is 30.

[0094] Then the above values can be substituted into the determination formula of a(y i ) to determine the inferred quantity a(y i ) of each subset of voting account features of the target non-leaf node. The calculation results are shown in Table 3:

[0095] Table 3

[0096] IP City = Wuhan IP City = Beijing IP City = Shanghai Brand 1 20-(30-21)*20 / 30＝14 15 10 Brand 2 10-(30-21)*10 / 30＝7 25 20

[0097] Assume that the target non - leaf node is Wuhan x3, then:

[0098]

[0099] Similarly, by traversing all target non - leaf nodes in the voting account feature tree, the vote - rigging suspicion scores of other target non - leaf nodes can be determined.

[0100] In this step, by determining the vote - rigging suspicion scores of all target non - leaf nodes in the voting account feature tree, and then judging whether there is vote - rigging behavior of the corresponding target anchor through the vote - rigging suspicion scores.

[0101] S113, identify the target anchor based on the vote - rigging suspicion score.

[0102] After the vote - rigging suspicion scores of all target non - leaf nodes are determined, identify the target anchor based on the vote - rigging suspicion scores.

[0103] In an optional embodiment, identifying the anchor based on the vote - rigging suspicion score includes:

[0104] Obtain the vote - rigging suspicion scores of all target non - leaf nodes;

[0105] If it is determined that there is any vote - rigging suspicion score greater than the detection threshold, determine that the target anchor is an anchor with vote - rigging behavior.

[0106] In addition, determine the target non - leaf nodes with vote - rigging suspicion scores greater than the detection threshold as abnormal nodes, and add abnormal marks to the abnormal nodes to improve the subsequent identification accuracy.

[0107] In this step, identifying the anchor according to the vote - rigging suspicion score, this identification process is relatively objective and lacks human subjective factors, so the identification accuracy can be ensured.

[0108] Based on the same inventive concept, this embodiment also provides a device for identifying an anchor with vote - rigging behavior, as Figure 3 shown, the device includes:

[0109] The first determination unit 31 is used to determine the voting account feature tree according to the voting account feature information of the target anchor; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information;

[0110] An obtaining unit 32, configured to obtain each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information;

[0111] A second determining unit 33, configured to determine a ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset for the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes;

[0112] An identifying unit 34, configured to identify the target anchor based on the ticket-scalping suspicion score.

[0113] For the specific functions of the above units, reference may be made to the corresponding descriptions in the foregoing method embodiments, and details are not described herein again. Since the device introduced in the embodiments of the present invention is the device adopted for implementing the method in the embodiments of the present invention, those skilled in the art can understand the specific structure and variations of the device based on the method introduced in the embodiments of the present invention, and details are not described herein again. Any device adopted by the method in the embodiments of the present invention falls within the scope of protection of the present invention.

[0114] The beneficial effects that the method, device, medium, and equipment for identifying an anchor with a ticket-scalping behavior provided by the present invention can bring are at least:

[0115] The present invention provides a method, device, medium, and equipment for identifying an anchor with a ticket-scalping behavior. The method includes: determining a voting account feature tree according to the voting account feature information of the target anchor; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information; obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information; for the target non-leaf node, determining a ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset; the target non-leaf node is any one of the non-leaf nodes; identifying the target anchor based on the ticket-scalping suspicion score. In this way, first, the voting account feature information is selected as the basic identification data. Since the above information is closely related to the actual ticket-scalping behavior, the accuracy of the basic identification data can be ensured; all possible voting account feature subsets are included in the voting account feature tree. When determining the ticket-scalping suspicion score according to the voting account feature subset, comprehensiveness and objectivity can be ensured, the accuracy of the ticket-scalping suspicion score can be ensured, and further, the identification accuracy of an anchor with a ticket-scalping behavior can be improved.

[0116] This embodiment provides a computer device 400, such as Figure 4As shown in the figure, it includes a memory 410, a processor 420, and a computer program 411 stored in the memory 410 and executable on the processor 420. When the processor 420 executes the computer program 411, the following steps are implemented:

[0117] Determine a voting account feature tree according to the voting account feature information of the target host; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information;

[0118] Obtain each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information;

[0119] For a target non-leaf node, based on the voting account feature subset, determine the ticket-scalping suspicion score corresponding to the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes;

[0120] Identify the target host based on the ticket-scalping suspicion score.

[0121] In a specific implementation process, when the processor 420 executes the computer program 411, any implementation manner in the foregoing embodiments can be implemented.

[0122] Since the computer device introduced in this embodiment is the device used to implement a method for identifying a host with a ticket-scalping behavior in an embodiment of the present application, based on the method introduced in the foregoing embodiments of the present application, those skilled in the art can understand the specific implementation manner and various variation forms of the computer device in this embodiment. Therefore, the specific implementation of how this server implements the method in the embodiments of the present application will not be described in detail here. As long as the device used by those skilled in the art to implement the method in the embodiments of the present application belongs to the scope protected by the present application.

[0123] Based on the same inventive concept, this embodiment provides a computer-readable storage medium 500, as Figure 5 shown, on which a computer program 511 is stored. When the computer program 511 is executed by a processor, the following steps are implemented:

[0124] Determine a voting account feature tree according to the voting account feature information of the target host; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information;

[0125] Obtain each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information;

[0126] For a target non-leaf node, based on the subset of voting account features, determine the vote rigging suspicion score corresponding to the target non-leaf node; the target non-leaf node is any one of the non-leaf nodes;

[0127] Identify the target host based on the vote rigging suspicion score.

[0128] In a specific implementation process, when the computer program 511 is executed by a processor, any implementation manner in the foregoing embodiments can be implemented.

[0129] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0130] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the specified functions in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0131] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the specified functions in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0132] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide means for implementing the specified functions in one process Figure 1 one process or multiple processes and / or blocks Figure 1Steps of functions specified in one or more boxes.

[0133] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0134] As described above, it is only the preferred embodiment of the present invention, and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for identifying a live streamer with vote rigging behavior, characterized in that, The method includes: Determining a voting account feature tree according to the voting account feature information of the target host; the voting account feature information includes: device attribute information, device IP information, and voting user attribute information; Obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node; the non-leaf nodes include: device attribute information, device IP information, or voting user attribute information; For a target non-leaf node, determining a ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset; the target non-leaf node is any one of the non-leaf nodes; Identifying the target host based on the ticket-scalping suspicion score; The determining a voting account feature tree according to the voting account feature information of the target host includes: Generating a root node of the voting account feature tree, and the root node is the parent node of the first-layer nodes; Obtaining the first feature information included in the device attribute information, the first feature value corresponding to the first feature information, the second feature information included in the device IP information, the second feature value corresponding to the second feature information, the third feature information included in the voting user attribute information, and the third feature value corresponding to the third feature information; Traversing the first feature value, the second feature value, and the third feature value, forming different-level feature value sets according to the first feature value, the second feature value, and the third feature value, and determining the corresponding layer nodes of the voting account feature tree according to the different-level feature value sets; wherein, The number of feature values included in the current layer node of the voting account feature tree is greater than the number of feature values included in the node of the previous layer of the current layer node, the current layer node is the second-layer node, and the node of the previous layer of the current layer node is the first-layer node; The determining a ticket-scalping suspicion score corresponding to the target non-leaf node based on the voting account feature subset includes: According to the formula determine the ballot stuffing suspicion score corresponding to the target non-leaf node ; where where x is the target non-leaf node, and is the subset of voting account features, i is the serial number of the subset of voting account features, n is the total number of subsets of voting account features, and is the actual number of the subset of voting account features, and is the predicted number of the subset of voting account features, and is the inferred number of the subset of voting account features; Wherein, by determining each reference host with the same level as the target host, obtaining the historical account number of each reference host that meets the voting account feature subset in the historical voting activity, and using the historical account number as the predicted number of the voting account feature subset of the target host; The actual number is the actual account number that participates in the voting activity of the target host and meets the voting account feature subset; The inferred number is the account number that meets the voting account feature subset inferred according to the parent node of the voting account feature subset; The obtaining each non-leaf node of the voting account feature tree and the voting account feature subset of each non-leaf node includes: Taking the first-layer nodes of the voting account feature tree as non-leaf nodes; Taking the nodes other than the first-layer nodes as the voting account feature subsets of each non-leaf node.

2. The method according to claim 1, wherein When the voting account feature subset is the subset corresponding to the target non-leaf node, the method further includes: According to the formula determine the inference quantity of the subset of the voting account features ; where The is the predicted quantity of the subset of voting account features, the is the subset of voting account features, where i is the serial number of the subset of voting account features, the is the predicted quantity of the non - leaf node, the is the actual quantity of the non - leaf node, and x is the target non - leaf node.

3. The method according to claim 1, characterized in that, When the voting account feature subset does not belong to the subset of the target non-leaf node, the method further includes: According to the formula Determine the inference quantity of the subset of voting account features ; where, the is the predicted quantity of the subset of voting account features, the is the subset of voting account features, and i is the serial number of the subset of voting account features.

4. The method according to claim 1, wherein The identifying the target host based on the ticket-scalping suspicion score includes: Obtain the vote rigging suspicion scores of all target non-leaf nodes; If it is determined that there is any vote rigging suspicion score greater than the detection threshold, determine that the target anchor is an anchor with vote rigging behavior.

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

6. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method according to any one of claims 1 to 4.

Images