WO2019134284A1

WO2019134284A1 - Method and apparatus for recognizing user, and computer device

Info

Publication number: WO2019134284A1
Application number: PCT/CN2018/082163
Authority: WO
Inventors: 王璐; 陈少杰; 张文明
Original assignee: 武汉斗鱼网络科技有限公司
Priority date: 2018-01-08
Filing date: 2018-04-08
Publication date: 2019-07-11
Also published as: CN108184148B; CN108184148A

Abstract

Provided are a method and apparatus for recognizing a user, and a computer device. The method comprises: respectively marking different labels for normal users and abnormal users in user samples; determining the value of similarity between any two users in the user samples; constructing a K nearest neighbor graph for all the users in the user samples using a K nearest neighbor algorithm; determining, according to the value of similarity between any two users in the user sample, a weight value of an edge constituted by any two users in the K nearest neighbor graph; according to a preset value for K and the weight value of the edge constituted by any two users in the K nearest neighbor graph, determining a first weight sum of edges constituted by connecting each of the remaining users to the normal users in the K nearest neighbor graph and a second weight sum of edges constituted by connecting each of the remaining users to the abnormal users; determining the label of each of the remaining users according to the first weight sum and the second weight sum; and recognizing each of the users according to the labels of the users.

Description

Method, device and computer device for identifying users

Technical field

The invention belongs to the field of live broadcast technology, and in particular relates to a method, a device and a computer device for identifying a user.

Background technique

With the development of the live broadcast platform, more and more users send barrage messages on the live broadcast platform. In order to increase the popularity of the live broadcast, some anchors will use the machine equipment to send a large amount of barrage information in the live broadcast room, which seriously affects the order of the live broadcast platform.

Generally speaking, most of the contents of the machine barrage are not different from the normal barrage information. The machine will imitate the barrage text of the normal user, so it is impossible to identify the abnormal user who sent the barrage, resulting in the order of the live platform being unavailable. Guarantee.

Summary of the invention

The present invention provides a method, a device, and a computer device for identifying a user, which are used to solve the problem that the abnormal user information of the broadcast screen in the live broadcast platform cannot be identified in the prior art, resulting in the problem of the prior art. The technical problem of the order of the live broadcast platform is not guaranteed.

An embodiment of the present invention provides a method for identifying a user, which is applied to a live broadcast platform, where the method includes:

Labeling different users and abnormal users in the user sample respectively, the users in the user samples include: normal users, abnormal users, and remaining users;

Determining a similar value between any two users in the user sample according to the number of live broadcasts corresponding to each user behavior, the device information used, and the Internet Protocol IP information that each user logs in in the live broadcast platform;

Constructing a K-nearest neighbor graph for all users in the user sample using the K-nearest neighbor algorithm;

Determining, according to a similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-nearest neighbor graph;

Determining, according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph, a first weight of an edge formed by each remaining user in the K-neighbor graph connected to the normal user And a second weight sum of edges of each remaining user connected to the abnormal user;

Determining a label of each remaining user according to the first weight sum and the second weight sum;

Each user is identified based on the user's tag.

In the foregoing solution, the normal user and the abnormal user in the user sample are respectively marked with different labels, including:

Obtaining a behavior log record and a user level of each user in the user sample in the live broadcast platform;

If the user's behavior log record is normal or the user level is a high-level user, then the user is determined to be a normal user;

Marking the first user for the normal user.

Obtaining a login account of each user in the live broadcast platform and a device identification ID corresponding to the login account;

If the login account corresponds to multiple identical device IDs, determine that the user corresponding to the login account is an abnormal user;

Marking the second user with the abnormal user.

In the foregoing solution, the information about the number of live broadcasts generated by each user behavior, the device information used, and the login IP information of each user in the live broadcast platform determine the similarity between any two users in the user sample. Values, including:

According to the formula

Determining the similarity value s(u,v) between any user u and user v;

When it is required to identify the user who sends the bullet information, the Ru is the number of live broadcasts that the user u sends the barrage within a preset time, and the R _v is the live room where the user v sends the barrage within the preset time. number; I _u is the user u of the user login information broadcast IP platform, the user v I _v is a user login information in the IP broadcast platform; the users u D _u is transmitted barrage The device information used at the time, the D _v is the device information used when the user v sends the barrage; the w _i (i=1, 2, 3, 4) is

The number of targets.

In the above solution, the determining, according to a preset K value and a weight value of an edge formed by any two users in the K-neighbor graph, determining that each remaining user in the K-neighbor graph is connected with the normal user The first weight sum of the edges, and the second weight sum of the edges formed by each remaining user connected to the abnormal user, including:

Determining, according to a preset K value, K edges formed by each remaining user in the K-neighbor graph and the normal user, and K edges formed by each remaining user connected to the abnormal user;

Calculating, according to weight values of edges of any two users in the K-neighbor graph, a first weight sum of K edges formed by each remaining user in the K-neighbor graph and the normal user, and each The second weight sum of the K edges formed by the remaining users connected to the abnormal user.

In the foregoing solution, determining, according to the similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-neighbor graph, including:

According to the formula

Determining a weight value of an edge formed by any two users in the K-nearest neighbor graph; wherein the s(u, v) is a similar value between any user u and user v, the a being a preset constant , a takes the value (0, 1).

In the above solution, the determining, according to the first weight and the second weight, a label of each remaining user, including:

Comparing the first weight sum and the second weight sum to determine a maximum weight sum;

The labels of each of the remaining users are marked as being consistent with the largest weight and the label of the corresponding user.

An embodiment of the present invention further provides an apparatus for identifying a user, where the apparatus includes:

a marking unit, configured to mark different labels for the normal user and the abnormal user in the user sample, where the users in the user sample include: a normal user, an abnormal user, and a remaining user;

a first determining unit, configured to determine, according to the quantity information of the live broadcast corresponding to each user behavior, the device information used, and the IP information that each user logs in in the live broadcast platform, determine between any two users in the user sample. Similar value

a building unit for constructing a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm;

a second determining unit, configured to determine, according to a similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-nearest neighbor graph;

a third determining unit, configured to determine, according to a preset K value, a first weight sum of edges of each remaining user in the K-neighbor graph connected to the normal user, and each remaining user and the abnormal user a second weight sum of the connected edges;

a fourth determining unit, configured to determine, according to the first weight sum and the second weight, a label of each remaining user;

The identification unit is configured to identify each user according to the user's label.

The present invention also provides a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the following steps:

Labeling different users for normal users and abnormal users in the user sample, the user samples include: normal users, abnormal users, and remaining users;

Determining a similarity value between any two users in the user sample according to the corresponding live broadcast quantity information, the used device information, and the IP information that each user logs in in the live broadcast platform;

Each user is identified based on the user's tag.

The invention also provides a computer device for identifying a user of a barrage, comprising:

At least one processor;

At least one memory communicatively coupled to the processor, wherein

The memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of the above.

The embodiment of the present invention provides a method, an apparatus, and a computer device for identifying a user. The method includes: marking different labels for a normal user and an abnormal user in a user sample, where the user in the user sample includes: a normal user, an abnormal user, and a remaining user; determining the number of live rooms corresponding to each user behavior, the device information used, and the IP information of each user logging in the live platform to determine any two users in the user sample. a similarity value between the two; constructing a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm; determining that any two users are in the K-nearest neighbor graph according to similarity values between any two users in the user sample a weight value of the edge; determining, according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph, each remaining user in the K-neighbor graph is connected with the normal user a first weight sum of the edges, and a second weight sum of edges of each remaining user connected to the abnormal user; according to the first weight sum a second weight and a label for each remaining user; each user is identified according to the user's tag; thus, different tags are marked for the normal user and the abnormal user in the user sample; since the remaining users are not yet identified Normal user and abnormal user, so K-nearest neighbor graph is constructed for all users in the user sample according to the K-nearest neighbor algorithm, and each remaining in the K-neighbor graph is determined according to the similarity value calculation between any two users in the user sample. a first weight sum of K edges formed by the user connected to the normal user, and a second weight sum of K edges formed by each remaining user connected to the abnormal user; and according to the first weight sum The second weight is determined and the label of each remaining user is determined. Finally, each user can be identified according to the label of each user; thus, the abnormal user in the live broadcast platform can be accurately identified, and the order of the live broadcast platform can be guaranteed.

DRAWINGS

Various other advantages and benefits will become apparent to those skilled in the art from a The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting. Throughout the drawings, the same reference numerals are used to refer to the same parts. In the drawing:

FIG. 1 is a schematic flowchart of a method for identifying a user according to an embodiment of the present invention;

2 is a schematic structural diagram of an apparatus for identifying a user according to an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of a computer device for identifying a user according to an embodiment of the present invention.

Detailed ways

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While the embodiments of the present invention have been shown in the drawings, the embodiments Rather, these embodiments are provided so that this disclosure will be more fully understood and the scope of the disclosure will be fully disclosed.

The technical solution of the present invention will be further described in detail below through the accompanying drawings and specific embodiments.

The embodiment provides a method for identifying a user, which is applied to a live broadcast platform. As shown in FIG. 1 , the method includes:

S110, marking different labels for normal users and abnormal users in the user sample.

In this step, the users in the user sample include: a normal user, an abnormal user, and a remaining user. The user sample may be a user that logs in within a preset time in the live broadcast platform. The user sample in this embodiment is a user sample that is counted daily in the live broadcast platform.

Here, the normal user and the abnormal user may be identified from the user sample by using a preset identification rule, and different labels are respectively marked for the normal user and the abnormal user in the user sample. However, the identification rule cannot identify all normal users and abnormal users in the user sample, and the identified normal users and abnormal users are only a part of users in the user sample, and thus there are some remaining users.

Specifically, when the normal user is identified and marked, the behavior log record and the user level of each user in the user sample in the live broadcast platform are obtained; if the user's behavior log record is normal or the user level is a high level The user is determined to be a normal user; the first user is marked with a first label, and the first label may be identified by a number or a letter or other, and is not limited herein. The first label in this embodiment is represented by a number, such as 1.

For example, when the user's behavior log record is used for identification, some users will have corresponding dot records in the behavior log record when sending the barrage or other trigger actions, and if there is a corresponding dot record in the log record, the corresponding trigger is also triggered. The action, such as sending a barrage message, indicates that the user is a normal user.

Here, when the user's level is identified, it is determined by the reference level 5, when the user level is greater than 5, the user is determined to be a high-level user; when the user level is less than 5, the user is determined to be a low-level user; For example, when it is determined that the user level is 10, the user is determined to be a high-level user; when the user level is 1 level, the user is determined to be a low-level user.

When the normal user is identified and marked, the login account of each user in the live broadcast platform and the device ID corresponding to the account are obtained; if the login account corresponds to multiple identical device IDs, then The user corresponding to the login account is an abnormal user; and the abnormal user is marked with a second tag. The second label may be identified by a number or a letter or other, and is not limited herein. The second label in this embodiment is represented by a number, such as zero.

For example, when sending the barrage information, multiple users use different devices to send the barrage information using the same account, and then determine that the user corresponding to the account is an abnormal user.

S111. Determine, according to the quantity information of the live broadcast corresponding to each user behavior, the device information used, and the login Internet Protocol IP information of each user in the live broadcast platform, determine a similarity value between any two users in the user sample. ;

In this step, the number of live broadcasts corresponding to each user behavior, the device information used, and the login Internet Protocol IP information of each user in the live broadcast platform determine between any two users in the user sample. Similar values. Here, each user generates a unique device identifier for each device by logging in to the live platform on the device.

Then you can determine the similarity between any two users in the user sample according to formula (1):

In formula (1), when it is necessary to identify the user who sends the bullet information, then the R _u is the number of live broadcasts that the user u sends the barrage within the preset time, and the R _v is the user v at the preset. number of live transmission time bomb curtain; I _u is the user u of the user login information broadcast IP platform, the user v I _v is a user login information in the IP broadcast platform; the D _u is the device information used when the user u sends the barrage, and the D _v is the device information used when the user v sends the barrage;

The N is the number of the feature indicators; the feature indicators related to the sending of the barrage may include: the number of times the barrage is transmitted within the preset time period, the interval between the transmission of the barrage, and the like. The device information is an ID of each device. When each device accesses the live broadcast platform, the live broadcast platform generates a unique ID for each device.

Here, when other types of users are required for identification, the above parameters may be adapted to other types of users. For example, other types of users may also include: users who send gifts, and the like.

S112. Construct a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm; determine weights of edges formed by any two users in the K-nearest neighbor graph according to similarity values between any two users in the user sample. value;

After determining the similarity value between any two users in the user sample, the K-nearest neighbor algorithm is used to construct a K-nearest neighbor graph for all users in the user sample. In the K-nearest neighbor graph, each user is equivalent to one node, two An edge is formed between nodes, and the similar value between users represents the relationship between nodes.

Then, the weight value of the edge formed by any two users in the K-nearest neighbor graph can be determined according to the similarity value between any two users in the user sample. The specific implementation method is shown in formula (2):

In the formula (2), the s(u, v) is a similar value between any user u and user v, the a is a preset constant, and a is a value of (0, 1).

It should be noted that when calculating the weight value of the constituent side between different users, the value of the a is the same.

S113. Determine, according to a preset K value and a weight value of an edge formed by any two users in the K-neighbor graph, a first edge of each side of the K-neighbor graph connected with the normal user. a sum of weights, and a second weight sum of edges of each remaining user connected to the abnormal user;

Determining the weight value of the edge formed by any two users in the K-nearest neighbor graph, determining the K according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph The first weight sum of the edges formed by each remaining user connected to the normal user in the neighbor graph, and the second weight sum of the edges formed by each remaining user connected to the abnormal user.

Specifically, if all the nodes are connected in the K-nearest neighbor graph, the K-nearest neighbor graph will be too complicated, so a K value is preset, and each remaining in the K-nearest neighbor graph is determined according to the preset K value. a K edge formed by the user connected to the normal user and a K edge formed by each remaining user connected to the abnormal user;

In this embodiment, the K value is set to 30, that is, each node is only connected to 30 nodes with the largest similar value, forming 30 edges.

For example, when the K value is 3, the remaining nodes include the A node, and the B, C, and D nodes have the largest similarity values; the edges of the A and B nodes have a weight of 0.4, and the edges of the A and C nodes The weight is 0.3, the weight of the edge formed by the A node and the D node is 0.1; the label of the B node and the D node is 1, and the label of the C node is 0; then the sum of the first weight is 0.5, and the sum of the second weight is 0.3. .

S114. Determine a label of each remaining user according to the first weight sum and the second weight sum;

After the first weight sum and the second weight sum are determined, a label of each remaining user is determined according to the first weight sum and the second weight sum.

Specifically, the first weight and the second weight sum are compared to determine a maximum weight sum; the label of each remaining user is marked as being consistent with the largest weight and the label of the corresponding user.

Taking the A node in the above steps as an example, since the sum of the first weights is 0.5, the sum of the second weights is 0.3. Then the label of the A node should be the same as the label of the normal user, that is, the label of the A node is 1.

In addition to the original node, other nodes are iteratively marked according to the above process until the label status of all nodes no longer changes, that is, the remaining users are marked. The original node refers to a node corresponding to an abnormal user in a user sample.

S115, each user is identified according to the user's label.

After tagging the remaining users, each user can be identified based on the user's tag. In this way, we can find out every possible abnormal user and ensure the order of the live broadcast platform.

Embodiment 2

Corresponding to the first embodiment, the embodiment provides a device for identifying a user. As shown in FIG. 2, the device includes: a marking unit 21, a first determining unit 22, a building unit 23, a second determining unit 24, and a third a determining unit 25, a fourth determining unit 26, and an identifying unit 27; wherein

The marking unit 21 is configured to respectively mark different labels for normal users and abnormal users in the user sample. The user in the user sample further includes: remaining users. The user sample may be a user that logs in within a preset time in the live broadcast platform. The user sample in this embodiment is a user sample that is counted daily in the live broadcast platform.

Here, the marking unit 21 may first identify a normal user and an abnormal user from the user sample by using a preset identification rule, and respectively mark different labels for the normal user and the abnormal user in the user sample. However, the identification rule cannot identify all normal users and abnormal users in the user sample, and the identified normal users and abnormal users are only a part of users in the user sample, and thus there are some remaining users.

Specifically, when the marking unit 21 identifies and marks a normal user, the behavior log record and the user level of each user in the user sample in the live broadcast platform are acquired; if the behavior log record of the user is normal or When the user level is a high-level user, the user is determined to be a normal user; the first user is marked with a first label, and the first label may be identified by a number or a letter or other, and is not limited herein. The first label in this embodiment is represented by a number, such as 1.

When the marking unit 21 identifies and marks the normal user, the login account of each user in the live sample platform and the device ID corresponding to the account are obtained; if the login account corresponds to multiple identical The device ID determines that the user corresponding to the login account is an abnormal user, and marks the second user with the abnormal user. The second label may be identified by a number or a letter or other, and is not limited herein. The second label in this embodiment is represented by a number, such as zero.

Then, the first determining unit 22 determines any two of the user samples according to the corresponding live broadcast quantity information, the used device information, and the login Internet Protocol IP information of each user in the live broadcast platform. Similar values between users. Here, each user generates a unique device identifier for each device by logging in to the live platform on the device.

Specifically, the first determining unit 22 determines a similar value between any two users in the user sample according to formula (1):

In formula (1), when it is necessary to identify the user who sends the bullet information, then the R _u is the number of live broadcasts that the user u sends the barrage within the preset time, and the R _v is the user v at the preset. number of live transmission time bomb curtain; I _u is the user u of the user login information broadcast IP platform, the user v I _v is a user login information in the IP broadcast platform; the D _u is the device information used when the user u sends the barrage, and the D _v is the device information used when the user v sends the barrage; the w _i (i=1, 2, 3, 4) is a weight coefficient, and

The x _ui is the i-th feature index related to the sending barrage of the user u, and the x _vi is the i-th feature index related to the sending barrage of the user v; the N is the number of the feature indicators; The feature indicators related to the sending of the barrage may include: the number of times the barrage is transmitted within the preset time period, the interval between the transmission of the barrage, and the like. The device information is an ID of each device. When each device accesses the live broadcast platform, the live broadcast platform generates a unique ID for each device.

After the first determining unit 22 determines the similarity value between any two users in the user sample, the constructing unit 23 is configured to construct a K-nearest neighbor graph for all users in the user sample by using the K-nearest neighbor algorithm, in the K-nearest neighbor Each user in the figure is equivalent to one node, and an edge is formed between two nodes. The similarity value between users represents the relationship between nodes.

Then, the second determining unit 24 may determine the weight value of the edge formed by any two users in the K-neighbor graph according to the similarity value between any two users in the user sample. The specific implementation method is shown in formula (2):

After the second determining unit 24 determines the weight value of the edge formed by any two users in the K-nearest neighbor graph, the third determining unit 25 is configured to use the preset K value and any two users according to the preset value. a weight value of an edge formed in the K-neighbor graph, determining a first weight sum of edges of each of the remaining neighbors in the K-neighbor graph and the normal user, and each remaining user is connected to the abnormal user The second weight of the side is.

For example, when the K value is 3, the remaining nodes include the A node, and the B, C, and D nodes with the largest similarity value; the edges of the A and B nodes have a weight of 0.4, and the A and C nodes form the edge. The weight is 0.3, the weight of the edge formed by the A node and the D node is 0.1; the label of the B node and the D node is 1, and the label of the C node is 0; then the sum of the first weight is 0.5, and the sum of the second weight is 0.3. .

After the first weight sum and the second weight sum are determined, the fourth determining unit 26 is configured to determine a label of each remaining user according to the first weight sum and the second weight sum.

After tagging the remaining users, the identification unit 27 can identify each user according to the user's tag. In this way, we can find out every possible abnormal user and ensure the order of the live broadcast platform.

Embodiment 3

The embodiment further provides a computer device for identifying a user. As shown in FIG. 3, the computer device includes: a radio frequency (RF) circuit 310, a memory 320, an input unit 330, a display unit 340, and an audio circuit. 350, WiFi module 360, processor 370, and power supply 380 and other components. Those skilled in the art will appreciate that the computer device architecture illustrated in FIG. 3 does not constitute a limitation to a computer device, and may include more or fewer components than those illustrated, or some components may be combined, or different component arrangements.

The specific components of the computer device will be specifically described below with reference to FIG. 3:

The RF circuit 310 can be used for receiving and transmitting signals, and in particular, receiving downlink information of the base station and processing it to the processor 370. Generally, RF circuit 310 includes, but is not limited to, at least one amplifier, transceiver, coupler, Low Noise Amplifier (LNA), duplexer, and the like.

The memory 320 can be used to store software programs and modules, and the processor 370 executes various functional applications and data processing of the computer devices by running software programs and modules stored in the memory 320. The memory 320 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application required for at least one function, and the like; the storage data area may store data created according to usage of the computer device, and the like. Further, the memory 320 may include a high speed random access memory, and may also include a nonvolatile memory such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The input unit 330 can be configured to receive input numeric or character information and to generate key signal inputs related to user settings and function control of the computer device. Specifically, the input unit 330 may include a touch panel 331 and other input devices 332. The touch panel 331 can collect input operations of the user and drive the corresponding connecting device according to a preset program. The touch panel 331 collects the output information and sends it to the processor 370. In addition to the touch panel 331, the input unit 330 may also include other input devices 332. Specifically, other input devices 332 may include, but are not limited to, one or more of a touch panel, function keys (such as volume control buttons, switch buttons, etc.), trackballs, mice, joysticks, and the like.

The display unit 340 can be used to display information input by the user or information provided to the user as well as various menus of the computer device. The display unit 340 can include a display panel 341. Alternatively, the display panel 341 can be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like. Further, the touch panel 331 can cover the display panel 341. When the touch panel 331 detects a touch operation on or near the touch panel 331, it transmits to the processor 370 to determine the type of the touch event, and then the processor 370 according to the input event. The type provides a corresponding visual output on display panel 341. Although the touch panel 331 and the display panel 341 are implemented as two separate components in FIG. 3 to implement input and input functions of the computer device, in some embodiments, the touch panel 331 may be integrated with the display panel 341. Implement the input and output functions of computer equipment.

An audio circuit 350, a speaker 351, and a microphone 352 can provide an audio interface between the user and the computer device. The audio circuit 350 can transmit the converted electrical data of the received audio data to the speaker 351, and convert it into a sound signal output by the speaker 351;

WiFi is a short-range wireless transmission technology. The computer device can help users to send and receive emails, browse web pages and access streaming media through the WiFi module 360. It provides users with wireless broadband Internet access. Although FIG. 3 shows the WiFi module 360, it can be understood that it does not belong to the essential configuration of the computer device, and may be omitted as needed within the scope of not changing the essence of the invention.

Processor 370 is a control center for computer devices that connects various portions of the entire computer device using various interfaces and lines, by running or executing software programs and/or modules stored in memory 320, and recalling data stored in memory 320. , performing various functions and processing data of the computer device, thereby performing overall monitoring of the computer device. Optionally, the processor 370 may include one or more processing units; preferably, the processor 370 may integrate an application processor, wherein the application processor mainly processes an operating system, a user interface, an application, and the like.

The computer device also includes a power source 380 (such as a power adapter) that supplies power to the various components. Preferably, the power source can be logically coupled to the processor 370 via a power management system.

The beneficial effects of the method, device and computer device for identifying a user provided by the embodiments of the present invention are at least:

The embodiment of the present invention provides a method, an apparatus, and a computer device for identifying a user. The method includes: marking different labels for a normal user and an abnormal user in a user sample, where the user in the user sample includes: a normal user, an abnormal user, and a remaining user; determining, according to the number of live broadcasts corresponding to each user behavior, the device information used, and the login Internet Protocol IP information of each user in the live broadcast platform, determining any two of the user samples. a similarity value between users; constructing a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm; determining any two users in the K-nearest neighbor graph according to similar values between any two users in the user sample The weight value of the edge formed in the middle; determining, according to the preset K value and the weight value of the edge formed by any two users in the K-nearest neighbor graph, each remaining user in the K-neighbor graph is connected to the normal user a first weight sum of the constituent edges, and a second weight sum of edges of each remaining user connected to the abnormal user; according to the first Reconciling the second weight with the label of each remaining user; identifying each user according to the user's tag; thus, marking different tags for the normal user and the abnormal user in the user sample; The normal user and the abnormal user of the remaining users are out, so the K-nearest neighbor graph is constructed for all users in the user sample according to the K-nearest neighbor algorithm, and the K-nearest neighbor is determined according to the similarity value calculation between any two users in the user sample. a first weight sum of K edges formed by each remaining user connected to the normal user, and a second weight sum of K edges formed by each remaining user connected to the abnormal user; a weight sum and the second weight and determining a label of each remaining user, and finally each user can be identified according to the label of each user; thus, the abnormal user in the live broadcast platform can be accurately identified, and the live broadcast is guaranteed. The order of the platform.

The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general purpose systems can also be used with the teaching based on the teachings herein. The structure required to construct such a system is apparent from the above description. Moreover, the invention is not directed to any particular programming language. It is to be understood that the invention may be embodied in a variety of programming language, and the description of the specific language has been described above in order to disclose the preferred embodiments of the invention.

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

Similarly, the various features of the invention are sometimes grouped together into a single embodiment, in the above description of the exemplary embodiments of the invention, Figure, or a description of it. However, the method disclosed is not to be interpreted as reflecting the intention that the claimed invention requires more features than those recited in the claims. Rather, as the following claims reflect, inventive aspects reside in less than all features of the single embodiments disclosed herein. Therefore, the claims following the specific embodiments are hereby explicitly incorporated into the embodiments, and each of the claims as a separate embodiment of the invention.

Those skilled in the art will appreciate that the modules in the devices of the embodiments can be adaptively changed and placed in one or more devices different from the embodiment. The modules or units or components of the embodiments may be combined into one module or unit or component, and further they may be divided into a plurality of sub-modules or sub-units or sub-components. In addition to such features and/or at least some of the processes or units being mutually exclusive, any combination of the features disclosed in the specification, including the accompanying claims, the abstract and the drawings, and any methods so disclosed, or All processes or units of the device are combined. Each feature disclosed in this specification (including the accompanying claims, the abstract and the drawings) may be replaced by alternative features that provide the same, equivalent or similar purpose.

In addition, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments and not other features, combinations of features of different embodiments are intended to be within the scope of the present invention. And different embodiments are formed. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

The various component embodiments of the present invention may be implemented in hardware, or in a software module running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or digital signal processing (DSP) may be used in practice to implement some of the gateways, proxy servers, some or all of the components in accordance with embodiments of the present invention. Or all features. The invention can also be implemented as a device or device program (e.g., a computer program and a computer program product) for performing some or all of the methods described herein. Such a program implementing the present invention may be stored on a computer readable storage medium or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form; when executed by the processor, the program implements the following steps: respectively marking different users and abnormal users in the user sample. a label, the user in the user sample includes: a normal user, an abnormal user, and a remaining user; a corresponding live broadcast quantity information, device information used, and a login Internet of each user in the live broadcast platform according to each user behavior The protocol IP information determines a similarity value between any two users in the user sample; constructs a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm; according to similar values between any two users in the user sample Determining a weight value of an edge formed by any two users in the K-nearest neighbor graph; determining the K-nearest neighbor graph according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph a first weight sum of edges of each of the remaining users connected to the normal user, and each remaining user and the different A second side connected to the user right weight and configuration; and in accordance with the first weight and the second weight is determined and each of the remaining user tag; for each user based on the user identification tag.

It is to be noted that the above-described embodiments are illustrative of the invention and are not intended to be limiting, and that the invention may be devised without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as a limitation. The word "comprising" does not exclude the presence of the elements or steps that are not recited in the claims. The word "a" or "an" The invention can be implemented by means of hardware comprising several distinct elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means can be embodied by the same hardware item. The use of the words first, second, and third does not indicate any order. These words can be interpreted as names.

The above is only the preferred embodiment of the present invention, and is not intended to limit the 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 Within the scope of protection of the present invention.

Claims

A method for identifying a user, wherein the application is in a live broadcast platform, the method comprising:

Labeling different users and abnormal users in the user sample respectively, the users in the user samples include: normal users, abnormal users, and remaining users;

Determining a similar value between any two users in the user sample according to the number of live broadcasts corresponding to each user behavior, the device information used, and the Internet Protocol IP information that each user logs in in the live broadcast platform;

Constructing a K-nearest neighbor graph for all users in the user sample using the K-nearest neighbor algorithm;

Determining, according to a similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-nearest neighbor graph;

Determining, according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph, a first weight of an edge formed by each remaining user in the K-neighbor graph connected to the normal user And a second weight sum of edges of each remaining user connected to the abnormal user;

Determining a label of each remaining user according to the first weight sum and the second weight sum;

Each user is identified based on the user's tag.
The method according to claim 1, wherein the labeling different labels for the normal user and the abnormal user in the user sample respectively comprises:

Obtaining a behavior log record and a user level of each user in the user sample in the live broadcast platform;

If the user's behavior log record is normal or the user level is a high-level user, then the user is determined to be a normal user;

Marking the first user for the normal user.
The method according to claim 1, wherein the labeling different labels for the normal user and the abnormal user in the user sample respectively comprises:

Obtaining a login account of each user in the live broadcast platform and a device identification ID corresponding to the login account;

If the login account corresponds to multiple identical device IDs, determine that the user corresponding to the login account is an abnormal user;

Marking the second user with the abnormal user.
The method according to claim 1, wherein the determining the user sample according to the number of live broadcasts generated by each user behavior, the device information used, and the login IP information of each user in the live broadcast platform Similar values between any two users, including:

According to the formula
Determining the similarity value s(u,v) between any user u and user v;

When it is necessary to identify the user who sends the bullet information, the R u is the number of live broadcasts that the user u sends the barrage within the preset time, and the R v is the live broadcast of the user v transmitting the barrage within the preset time. The number is the number of times; the I u is the login IP information of the user u user in the live broadcast platform, and the I v is the login IP information of the user v user in the live broadcast platform; the D u is the user u sending the bullet The device information used in the screen, the D v is the device information used when the user v sends the barrage; the w i (i=1, 2, 3, 4) is
The number of targets.
The method according to claim 1, wherein said determining each remaining in said K-neighbor graph according to a preset K value and a weight value of an edge formed by any two users in said K-nearest neighbor graph a first weight sum of edges formed by the user connected to the normal user, and a second weight sum of edges formed by each remaining user connected to the abnormal user, including:

Determining, according to a preset K value, K edges formed by each remaining user in the K-neighbor graph and the normal user, and K edges formed by each remaining user connected to the abnormal user;

Calculating, according to weight values of edges of any two users in the K-neighbor graph, a first weight sum of K edges formed by each remaining user in the K-neighbor graph and the normal user, and each The second weight sum of the K edges formed by the remaining users connected to the abnormal user.
The method according to claim 1, wherein said determining, according to a similarity value between any two users in said user sample, weight values of edges of any two users in said K-neighbor graph, including :

According to the formula
Determining a weight value of an edge formed by any two users in the K-nearest neighbor graph; wherein, s(u, v) is a similar value between any user u and user v, and the a is a preset constant , a takes the value (0, 1).
The method according to claim 1, wherein said determining a label of each remaining user based on said first weight sum and said second weight sum comprises:

Comparing the first weight sum and the second weight sum to determine a maximum weight sum;

The labels of each of the remaining users are marked as being consistent with the largest weight and the label of the corresponding user.
A device for identifying a user, the device comprising:

a marking unit, configured to mark different labels for the normal user and the abnormal user in the user sample, where the users in the user sample include: a normal user, an abnormal user, and a remaining user;

a first determining unit, configured to determine, according to the quantity information of the live broadcast corresponding to each user behavior, the device information used, and the IP information that each user logs in in the live broadcast platform, determine between any two users in the user sample. Similar value

a building unit for constructing a K-nearest neighbor graph for all users in the user sample by using a K-nearest neighbor algorithm;

a second determining unit, configured to determine, according to a similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-nearest neighbor graph;

a third determining unit, configured to determine, according to a preset K value, a first weight sum of edges of each remaining user in the K-neighbor graph connected to the normal user, and each remaining user and the abnormal user a second weight sum of the connected edges;

a fourth determining unit, configured to determine, according to the first weight sum and the second weight, a label of each remaining user;

The identification unit is configured to identify each user according to the user's label.
A computer readable storage medium having stored thereon a computer program, wherein the program, when executed by the processor, implements the following steps:

Labeling different users and abnormal users in the user sample respectively, the users in the user samples include: normal users, abnormal users, and remaining users;

Determining a similarity value between any two users in the user sample according to the corresponding live broadcast quantity information, the used device information, and the IP information that each user logs in in the live broadcast platform;

Constructing a K-nearest neighbor graph for all users in the user sample using the K-nearest neighbor algorithm;

Determining, according to a similarity value between any two users in the user sample, a weight value of an edge formed by any two users in the K-nearest neighbor graph;

Determining, according to a preset K value and a weight value of an edge formed by any two users in the K-nearest neighbor graph, a first weight of an edge formed by each remaining user in the K-neighbor graph connected to the normal user And a second weight sum of edges of each remaining user connected to the abnormal user;

Determining a label of each remaining user according to the first weight sum and the second weight sum;

Each user is identified based on the user's tag.
A computer device for identifying a user of a barrage, comprising:

At least one processor;

At least one memory communicatively coupled to the processor, wherein

The memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1-7.