Method and system for generating audience tags for target accounts

Abstract

The method and system for generating an audience label of a target account provided in the specification obtain a target label from a plurality of known labels corresponding to all users of a target platform, the target label corresponding to a target group of users in the all users, calculate a group tendency value of the target group of users with the determined label to the target account, and then generate an audience label of the target account based on the target label. The target group of users in the specification has a determined label. From the perspective of the determined label, the group tendency value is determined, so that the target group of users has stronger explainability and larger operation space.

Description

TECHNICAL FIELD

[0001] The present specification relates to the field of Internet, and in particular, to a method and system for generating audience label for target account. BACKGROUND

[0002] Some application platforms can set labels for user accounts according to the characteristics of the user accounts, so as to facilitate account operation by operation personnel. For example, in the Alipay platform, various users can share videos, images and / or texts through their accounts in the life channel, and the accounts can be given labels such as food bloggers and makeup bloggers according to the shared content. For another example, in a shopping platform, a merchant user can display product information through a merchant account, and the merchant account can be given a label according to the displayed product information.

[0003] Taking live streaming by an anchor through an account as an example, the audience groups of different types of live streaming videos can be different. In order to facilitate operation, the audience group of the account can be determined, and the label of the audience group can be taken as one of the understanding labels of the account. In the prior art, when generating an audience label for an account, the preference value of each user on the application platform for the account is usually calculated one by one, the preference value reflects the degree of love of a single user for the content of the account, and then the users with a preference value greater than a threshold value are grouped into a group, and the group is the audience group of the account. However, it is difficult to summarize the common characteristics of the audience group, and it is impossible to determine from the audience group which users with what characteristics / labels prefer the content of the account, resulting in poor explainability of the audience group and small operation space. SUMMARY

[0004] The method and system for generating an audience label for a target account provided by the present specification have stronger explainability and larger operation space.

[0005] In a first aspect, the present specification provides a method for generating an audience label for a target account, comprising: obtaining a target label from a plurality of known labels corresponding to all users of a target platform, the target label corresponding to a target group of the all users; determining a group tendency value of the target group for the target account, the target account providing a target service to the target group, the group tendency value reflecting a degree of love of the target group for the target account; and generating the audience label of the target account based on the target label.

[0006] In some embodiments, the target service comprises at least one of the following: online video, offline video, audio, image, text and product display.

[0007] In some embodiments, the determining the crowd tendency value of the target account for the target crowd comprises: determining a comprehensive crowd index of the target crowd interacting with the target account through the target service in at least one interaction dimension, each interaction dimension corresponding to at least one interaction behavior between the target crowd and the target account; determining a confidence of the crowd tendency value, the confidence reflecting a reliability of the crowd tendency value; and determining the crowd tendency value based on the comprehensive crowd index and the confidence.

[0008] In some embodiments, the determining the crowd tendency value based on the comprehensive crowd index and the confidence comprises: determining the crowd tendency value by combining the comprehensive crowd index and the confidence through an activation function.

[0009] In some embodiments, the determining the comprehensive crowd index comprises: determining a single-dimensional crowd index of the target crowd interacting with the target account in each interaction dimension respectively; obtaining a weight corresponding to each interaction dimension, the weight reflecting an interaction depth embodied by the at least one interaction behavior corresponding thereto; and performing weighted summation on each single-dimensional crowd index and the corresponding weight to obtain the comprehensive crowd index.

[0010] In some embodiments, the determining the single-dimensional crowd index of the target crowd interacting with the target account in each interaction dimension respectively comprises, for each interaction dimension respectively: determining a first quantity proportion of the target crowd in a first user quantity on the target platform, the first user quantity being a user quantity that has the at least one interaction behavior corresponding to the interaction dimension with the target account within a first preset time period; determining a second quantity proportion of the target crowd in a second user quantity on the target platform, the second user quantity being a user quantity that has the at least one interaction behavior corresponding to the interaction dimension with any homogeneous account within the first preset time period, the homogeneous account being an account on the target platform that has the same behavior and / or operating nature as the target account; and determining a ratio of the first quantity proportion to the second quantity proportion as the single-dimensional crowd index in the interaction dimension.

[0011] In some embodiments, the interaction dimension at least comprises one of the following: a watching dimension, a clicking dimension, a following dimension, an interacting dimension, and a participating activity dimension, and the interaction behavior at least comprises one of the following: watching, clicking, following, interacting, and participating in an activity.

[0012] In some embodiments, the weights corresponding to the watching dimension, the clicking dimension, the following dimension, the interacting dimension, and the participating activity dimension are sequentially increased.

[0013] In some embodiments, the confidence of the crowd tendency value is positively correlated with a historical interaction number of the target account in the second preset time period.

[0014] In some embodiments, the determining the confidence of the crowd tendency value comprises: obtaining a historical interaction number of the target account in the second preset time period in at least one interaction behavior; obtaining an adjustment factor; and determining the confidence based on the historical interaction number and the adjustment factor.

[0015] In some embodiments, the activation function is configured to normalize the crowd tendency value.

[0016] In some embodiments, the generating the account tag of the target account based on the target tag comprises: determining that the crowd tendency value is greater than or equal to a preset threshold, and taking the target tag as the audience tag of the target account.

[0017] In some embodiments, the obtaining the target tag from a plurality of known tags corresponding to a total number of users of a target platform comprises: obtaining a plurality of crowd portraits corresponding to a plurality of crowds from the target platform, the plurality of crowds comprising a plurality of access users constituting the total number of users, each of the plurality of crowd portraits corresponding to a crowd tag, and each of the crowd tags comprising at least one known tag; and obtaining the target tag from the plurality of crowd tags corresponding to the plurality of crowd portraits.

[0018] In some embodiments, further comprising: obtaining the audience tag of the target account; obtaining a plurality of first access users meeting the audience tag; and delivering the target service of the target account to an account of each of the first access users.

[0019] In some embodiments, further comprising: obtaining a distribution model trained by a plurality of understanding tags of a plurality of registered accounts of the target platform, each of the understanding tags comprising a plurality of tags for identifying characteristics of the registered accounts, and each of the understanding tags comprising the audience tag; predicting a plurality of second access users meeting the understanding tag of the target account by using the distribution model; and delivering the target service of the target account to an account of each of the second access users.

[0020] In a second aspect, the present specification also provides a system for generating a crowd label for a target account, comprising at least one storage medium and at least one processor, the at least one storage medium storing at least one set of instructions for generating a crowd label for a target account; the at least one processor is in communication connection with the at least one storage medium, wherein when the system for generating a crowd label for a target account is running, the at least one processor reads the at least one set of instructions and implements the method for generating a crowd label for a target account according to the first aspect of the present specification.

[0021] From the above technical solutions, the method and system for generating an audience label for a target account provided by the present specification obtain a target label from a plurality of known labels corresponding to all users of a target platform, the target label corresponds to a target crowd in the all users, calculate a crowd tendency value of the target crowd with the determined label to the target account, and then generate an audience label for the target account based on the target label. The target crowd of the present specification has a determined label, and the crowd tendency value is determined from the perspective of the determined label, so that the target crowd has stronger interpretability and larger operation space.

[0022] Other functions of the method and system for generating an audience label for a target account provided by the present specification will be partially listed in the following description. According to the description, the following numbers and examples will be apparent to those of ordinary skill in the art. The creative aspects of the method and system for generating an audience label for a target account provided by the present specification can be fully explained by practicing or using the methods, devices and combinations described in the following detailed examples. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present specification, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present specification, and those of ordinary skill in the art can also obtain other drawings according to these drawings without creative labor.

[0024] Figure 1 An application scenario schematic diagram of a system 001 for generating an audience label for a target account according to some embodiments of the present specification is shown;

[0025] Figure 2 A hardware structure diagram of a computing device 600 according to some embodiments of the present specification is shown;

[0026] Figure 3 A flowchart of a method P100 for generating an audience label for a target account according to some embodiments of the present specification is shown; and

[0027] Figure 4 A flowchart illustrating a method S140 of generating audience tags for a target account according to some embodiments of the present specification is shown. DETAILED DESCRIPTION

[0028] The following description provides specific applications and requirements of the present specification, which is intended to enable a person skilled in the art to manufacture and use the contents of the present specification. Various local modifications of the disclosed embodiments are obvious to those skilled in the art, and the general principles defined herein can be applied to other embodiments and applications without departing from the spirit and scope of the present specification. Therefore, the present specification is not limited to the embodiments shown, but is consistent with the widest scope of the claims.

[0029] The terms used herein are only for the purpose of describing specific example embodiments, and are not limiting. For example, unless the context clearly indicates otherwise, as used herein, the singular forms "a", "an", and "the" can also include the plural forms. When used in the present specification, the terms "comprise", "include" and / or "contain" mean that the associated whole, step, operation, element and / or component exists, but do not exclude the presence of one or more other features, whole, step, operation, element, component and / or group or other features, whole, step, operation, element, component and / or group can be added in the system / method.

[0030] In view of the following description, these features of the present specification and other features, the operation and function of related elements of the structure, and the economy of combination and manufacture of components can be obviously improved. Referring to the drawings, all of which form part of the present specification. However, it should be clearly understood that the drawings are for illustrative and descriptive purposes only and are not intended to limit the scope of the present specification. It should also be understood that the drawings are not drawn to scale.

[0031] The flowchart used in the present specification shows the operation of system implementation according to some embodiments of the present specification. It should be clearly understood that the operations of the flowchart can not be implemented in sequence. On the contrary, the operations can be implemented in reverse order or simultaneously. In addition, one or more other operations can be added to the flowchart. One or more operations can be removed from the flowchart.

[0032] For the convenience of description, the present specification explains the terms that will appear in the following description:

[0033] Life number host: a blogger / host who performs live streaming in the life number of Alipay.

[0034] Crowd tendency: the preference of the crowd for a target account (such as a host account).

[0035] TGI index: Target Group Index refers to the proportion of groups with a certain characteristic in the target group / the proportion of groups with the same characteristic in the overall population.

[0036] Figure 1 FIG. 1 shows a schematic diagram of an application scenario of a system 001 for generating audience tags for a target account according to some embodiments of this specification. Figure 1 As shown, the system 001 for generating audience tags for target accounts (hereinafter referred to as system 001) can be used in any scenario of generating account tags. Figure 1 As shown, the system 001 may include an access user 100 , a client 200 , a server 300 and a network 400 .

[0037] The accessing user 100 may be any one of all users accessing the target platform. When the accessing user 100 performs relevant operations on the target platform, the target platform may provide the accessing user 100 with a computer software and / or hardware operating environment.

[0038] The target platform can be run on the client 200, and the client 200 provides a running environment for the target platform. For example, the client 200 can display the user interface of the target platform through a webpage, or install an application corresponding to the target platform, or display a link of the target platform for the user to click. The access user 100 can access the target platform using the client 200, so that the client 200 can collect the user information of the access user 100. In some embodiments, the method of generating audience labels for a target account can be executed on the client 200. For example, the client 200 can analyze the used user information of the access user 100 to obtain one or more highly refined labels. The client 200 can also receive the user information of other access users 100 in the full user set except the used access user 100 from the server 300, and analyze the user information of the other access users 100 to obtain one or more highly refined labels, so that one or more known labels of each access user 100 in the full user set are obtained, that is, the full user corresponds to multiple known labels. Therefore, the client 200 executes the method of generating audience labels for a target account based on the multiple known labels of the full user. In another embodiment, the client 200 can also directly receive the multiple known labels corresponding to the full user from the server 300, that is, the analysis and conversion from the user information to the known labels are completed by the server 300. Therefore, the client 200 executes the method of generating audience labels for a target account based on the multiple known labels of the full user. At this time, the client 200 can store data or instructions for executing the method of generating audience labels for a target account described in this specification, and can execute or be used to execute the data or instructions. In some embodiments, the client 200 can include a hardware device with data information processing function and necessary programs required to drive the hardware device to work. For example, Figure 1As shown, the client 200 can be communicatively connected with the server 300. In some embodiments, the server 300 can be communicatively connected with a plurality of clients 200. In some embodiments, the client 200 can interact with the server 300 through the network 400 to receive or send messages, etc., such as receiving or sending known labels, target labels of full users. In some embodiments, the client 200 can include a mobile device, a tablet, a notebook, a built-in device of a motor vehicle, or the like, or any combination thereof. In some embodiments, the mobile device can include a smart home device, a smart mobile device, a virtual reality device, an augmented reality device, or the like, or any combination thereof. In some embodiments, the smart home device can include a smart television, a desktop computer, or the like, or any combination thereof. In some embodiments, the smart mobile device can include a smartphone, a personal digital assistant, a gaming device, a navigation device, or the like, or any combination thereof. In some embodiments, the virtual reality device or the augmented reality device can include a virtual reality headset, a virtual reality glasses, a virtual reality patch, an augmented reality headset, an augmented reality glasses, an augmented reality patch, or the like, or any combination thereof. For example, the virtual reality device or the augmented reality device can include Google glasses, a head-mounted display, VR, or the like. In some embodiments, the built-in device in the motor vehicle can include an on-board computer, an on-board television, or the like. In some embodiments, the client 200 can include an image acquisition device for acquiring real face images, such as face images of the access user 100. In some embodiments, the image acquisition device can be a two-dimensional image acquisition device (such as an RGB camera), or a two-dimensional image acquisition device (such as an RGB camera) and a depth image acquisition device (such as a 3D structured light camera, a laser detector, or the like). In some embodiments, the client 200 can be a device with positioning technology for positioning the location of the client 200.

[0039] In some embodiments, the client 200 can be installed with one or more applications (APPs). The APPs can provide the access user 100 with the ability and interface to interact with the outside world through the network 400. The APPs include, but are not limited to, web browser type APPs, search type APPs, chat type APPs, shopping type APPs, video type APPs, financial type APPs, instant messaging tools, email clients, social platform software, and the like. In some embodiments, the client 200 can be installed with a target APP corresponding to a target platform. The access user 100 can trigger an access request through the target APP. The target APP can receive the access request and allow the access user 100 to access. The target APP can determine or obtain the known tags of the access user 100 for the client 200, obtain target tags based on the known tags, and the target tags correspond to the target population in the full amount of users. The target APP can execute the method for generating audience tags for a target account based on the target tags.

[0040] The server 300 can be a server that provides various services, such as a background server that provides support for pages displayed on the client 200. In some embodiments, the method for generating audience tags for a target account can be executed on the server 300. For example, the access user 100 accesses the target APP through the client 200, and the client 200 can collect the user information of the access user 100 and send the user information to the server 300 through the network 400. The server 300 can analyze the user information of the access user 100 to obtain one or more highly refined known tags. In this way, one or more known tags of the access user 100 using other clients 200 can also be obtained, so that the known tags of each access user 100 in the full amount of users can be obtained. Thus, the server 300 executes the method for generating audience tags for a target account based on the multiple known tags of the full amount of users. At this time, the server 300 can store data or instructions for generating audience tags for a target account as described in this specification, and can execute or be used to execute the data or instructions. In some embodiments, the server 300 can include a hardware device with data information processing functions and the necessary programs to drive the hardware device to work. The server 300 can be in communication connection with multiple clients 200 and receive data sent by the client 200.

[0041] The network 400 is a medium for providing communication connection between the client 200 and the server 300. The network 400 can facilitate the exchange of information or data. For example, Figure 1As shown, the client 200 and the server 300 can be connected with the network 400, and transmit information or data to each other through the network 400. In some embodiments, the network 400 can be any type of wired or wireless network, or a combination thereof. For example, the network 400 can include a cable network, a wired network, a fiber optic network, a telecommunication network, an intranet, the Internet, a Local Area Network (LAN), a Wide Area Network (WAN), a wireless Local Area Network (WLAN), a Metropolitan Area Network (MAN), a Wide Area Network (WAN), a Public Switched Telephone Network (PSTN), a Bluetooth network, a ZigBee network, a Near Field Communication (NFC) network, or the like. In some embodiments, the network 400 can include one or more network access points. For example, the network 400 can include wired or wireless network access points, such as base stations or Internet exchange points, through which one or more components of the client 200 and the server 300 can connect to the network 400 to exchange data or information.

[0042] It should be understood that Figure 1 The number of the client 200, the server 300, and the network 400 is only illustrative. Any number of the client 200, the server 300, and the network 400 can be provided according to implementation needs.

[0043] It should be noted that the method for generating an audience tag for a target account can be completely executed on the client 200, completely executed on the server 300, partially executed on the client 200, or partially executed on the server 300.

[0044] Figure 2 A hardware structure diagram of a computing device 600 is shown, which is provided according to some embodiments of the present specification. The computing device 600 can execute the method for generating an audience tag for a target account described in the present specification. The method for generating an audience tag for a target account is described in other parts of the present specification. When the method for generating an audience tag for a target account is executed on the client 200, the computing device 600 can be the client 200. When the method for generating an audience tag for a target account is executed on the server 300, the computing device 600 can be the server 300. When the method for generating an audience tag for a target account can be partially executed on the client 200 and partially executed on the server 300, the computing device 600 can be the client 200 and the server 300.

[0045] As Figure 2 As shown, the computing device 600 can include at least one storage medium 630 and at least one processor 620. In some embodiments, the computing device 600 can further include a communication port 650 and an internal communication bus 610. Meanwhile, the computing device 600 can further include an I / O component 660.

[0046] Internal communication bus 610 can connect different system components, including storage medium 630, processor 620, and communication port 650.

[0047] I / O components 660 support input / output operations of computing device 600 and other components.

[0048] Communication port 650 is used for data communication between computing device 600 and the outside world. For example, communication port 650 can be used for data communication between computing device 600 and network 400. Communication port 650 can be a wired communication port or a wireless communication port.

[0049] Storage medium 630 can include a data storage device. The data storage device can be a non-transitory storage medium or a transitory storage medium. For example, the data storage device can include one or more of a disk 632, a read-only memory (ROM) 634, or a random access memory (RAM) 636. Storage medium 630 can store at least one instruction set for training an image detection network, and can also store at least one instruction set corresponding to the image detection network after training. The instructions are computer program codes, which can include programs, routines, objects, components, data structures, processes, modules, etc. that perform the method of face recognition provided in the specification.

[0050] The at least one processor 620 can be communicatively connected with the at least one storage medium 630 and the communication port 650 through the internal communication bus 610. The at least one processor 620 is configured to execute the at least one instruction set described above. When the computing device 600 is running, the at least one processor 620 can read the at least one instruction set for training the image detection network, and execute the training method of the image detection network provided in the present specification according to the instructions of the at least one instruction set. When the computing device 600 is running, the at least one processor 620 can also read the at least one instruction set corresponding to the image detection network, and execute the image detection method provided in the present specification according to the instructions of the at least one instruction set. The processor 620 can execute all steps of the training method of the image detection network and the image detection method. The processor 620 can be in the form of one or more processors, and in some embodiments, the processor 620 can include one or more hardware processors, such as a microcontroller, a microprocessor, a reduced instruction set computer (RISC), an application-specific integrated circuit (ASIC), an application-specific instruction set processor (ASIP), a central processing unit (CPU), a graphics processing unit (GPU), a physics processing unit (PPU), a microcontroller unit, a digital signal processor (DSP), a field programmable gate array (FPGA), an advanced RISC machine (ARM), a programmable logic device (PLD), any circuit or processor capable of executing one or more functions, or the like, or any combination thereof. For the sake of illustration only, only one processor 620 is described in the computing device 600 in the present specification. However, it should be noted that the computing device 600 in the present specification can also include multiple processors, and thus the operations and / or method steps disclosed in the present specification can be executed by one processor as described in the present specification, or jointly executed by multiple processors. For example, if the processor 620 of the computing device 600 in the present specification executes step A and step B, it should be understood that step A and step B can also be executed jointly or separately by two different processors 620 (e.g., a first processor executes step A, and a second processor executes step B, or the first and second processors jointly execute steps A and B).

[0051] Figure 3 A flowchart of a method P100 for generating audience tags for a target account is shown, according to some embodiments of the present specification. As described previously, the computing device 600 can execute the method P100 for generating audience tags for a target account described in the present specification. Specifically, the processor 620 can read the instruction set stored in its local storage medium, and then execute the method P100 for generating audience tags for a target account described in the present specification according to the provisions of the instruction set. As shown, the method P100 can include: Figure 3

[0052] ​S120: obtaining a target label from a plurality of known labels corresponding to all users of the target platform, the target label corresponding to a target group of users in the all users.

[0053] The target platform refers to a platform that provides a target service for the all users when they access the target platform, such as an Alipay platform, a music platform, a social platform, an instant messaging platform, a shopping platform, and the like. For ease of description, any user in the all users is referred to as an access user. The all users refer to a set of all users who have accessed the target platform. The access can be that the access user accesses the target platform without account registration on the target platform but in the identity of a visitor. The access can also be that the user has completed account registration on the target platform and accesses the target platform using the registered account to log in.

[0054] The target platform can be displayed to the access user in the form of a webpage, for example, the processor 620 can open the webpage of the Alipay in the browser for the access user to access. The target platform can be in the form of an application program for the access user to download and use, for example, the access user can download and install the Alipay APP (application) on the client, and then access the Alipay APP. The target platform can also be displayed to the access user in the form of a link, and the access user can click the link to enter the target platform, for example, the processor 620 can display the Alipay link on a certain instant messaging software, and the access user clicks the Alipay link on the instant messaging software to jump to the Alipay platform to access.

[0055] The target service provided by the target platform for the access user can be a video, an audio, an image, a text, a commodity display, and the like, wherein the video can be an online video (also referred to as a live video) or an offline video (also referred to as a recorded video). For example, the tab 3 (the third interface label) below the interface of the Alipay APP is a life channel, and the access user can watch the live video, the recorded video, the image, the text, and the like published by multiple accounts in the life channel, and the anchor account for live broadcast can be referred to as a life anchor; at the same time, the access user can also use his own registered account to publish live video, recorded video, image, text, and the like for others to watch. For example, on a shopping APP, multiple merchants display the commodity information of the sold commodities in the form of commodity links through their merchant accounts, and the access user can view the commodity information and click the commodity link to purchase.

[0056] When the access user accesses the target platform, the target platform can collect user information of the access user, and the processor 620 can analyze the user information to obtain one or more labels of the access user, so as to obtain the label of each access user in the full amount of users. For example, the label of the access user A is <female, 22 years old, college student, no income, history, running, skin care>, the label of the access user B is <male, 40 years old, state-owned enterprise management, high income, stock, fund, food>, and the label of the access user C is <female, 30 years old, private enterprise employee, medium-high income, home, baby, skin care>. After the processor 620 obtains the user information of the access user, since the user information is determined and known, the label obtained by the processor 620 according to the user information is also determined and known, such as the above examples of the access users A, B and C, and therefore the label of each access user in the full amount of users can be referred to as a known label.

[0057] After obtaining the known label of each access user, the processor 620 can construct a plurality of crowd portraits corresponding to the full amount of users for the target platform, and each crowd portrait refers to a portrait obtained by representing the characteristics of a crowd through a crowd label. Each crowd includes a plurality of access users, and each crowd label includes one or more known labels. For example, the crowd label of the crowd A is <female, 20-30 years old, skin care>, the crowd label of the crowd B is <male, 30-50 years old, investment>, and the crowd label of the crowd C is <female, 25-35 years old, medium-high income, skin care>.

[0058] In order to identify the preferences of different crowds for different accounts, the processor 620 can determine the preference of a corresponding crowd for a certain account according to the crowd label of each crowd portrait. For any crowd label, in some embodiments, the processor 620 can obtain a plurality of crowd portraits corresponding to a plurality of crowds from the target platform, the access users included in the plurality of crowds constitute the full amount of users, obtain a target label from the plurality of crowd labels corresponding to the plurality of crowd portraits, and the crowd represented by the target label is referred to as a target crowd. For example, the target label is <female, 25-35 years old, medium-high income, skin care>, and the target crowd is the crowd C. Of course, the present specification can also obtain the target label by other methods, for example, a target label is set in advance according to a given task, and access users meeting the target label are found from the full amount of users to form a target crowd. For example, the given task is to determine the degree of love of a female between 25 and 35 years old, with medium-high income and love for skin care, for the anchor account A, and the target label can be set in advance as <female, 25-35 years old, medium-high income, skin care>, and then the target crowd is found from the full amount of users. The present specification does not make specific limitations on the way of obtaining the target label.

[0059] After the processor 620 obtains the target label, the method P100 can further include:

[0060] S140: determining a crowd tendency value of the target crowd to the target account, the target account providing a target service to the target crowd, and the crowd tendency value reflecting a degree of favor of the target crowd to the target account.

[0061] The target account can be a registered account on a target platform, and the target platform can provide the target crowd with a target service such as an online video, an offline video, an audio, an image, a text, a commodity display, and the like through the target account, and the target crowd can interact with the target account through the target service. The interaction can include multiple dimensions, such as a viewing dimension, a clicking dimension, a following dimension, an interaction dimension, and a participating activity dimension, and the like. Accordingly, the interaction behavior can include viewing, clicking, following, interacting, and participating in activities. Of course, the interaction can also refer to other interaction behaviors, such as listening, collecting, liking, commenting, and the like, which are not limited in the embodiments of the present specification. For example, in the life channel of Alipay, the host can publish a live video (online video) through the host account, and the target crowd can watch the live video, click the icon in the live video, follow the host account, publish comments in the live video, send virtual gifts to the host, have a voice chat with the host, and participate in marketing activities in the live video. For example, in a shopping platform, the host can publish a live video with a shopping link through the host account, and the target crowd can watch the live video, click the red envelope icon in the live video, like the host by continuously clicking the screen, view virtual goods by clicking the shopping bag icon, add virtual goods to the shopping cart, follow the host account, collect the store, have a text interaction with the host, have a voice interaction with the host, and participate in the live activity in the live room. For example, in an audio platform, a singer can publish a song through a music account, and the target crowd can interact with the music account by listening to the song and collecting the music account. For example, in an instant messaging platform, the owner of a public account publishes text and image content through the public account, and the target crowd can interact with the public account by reading, collecting, liking, and commenting on the text and image information.

[0062] Figure 4 A flowchart of a method S140 for generating an audience label for a target account is shown according to some embodiments of the present specification. In some embodiments, step S140 can include steps S142-S146:

[0063] S142: determining a comprehensive crowd index of the target crowd interacting with the target account through the target service in one or more dimensions.

[0064] In some embodiments, the processor 620 can determine a single-dimension group index of the target population interacting with the target account under each interaction dimension, such as a single-dimension TGI. Specifically, for each interaction dimension respectively: the processor 620 can determine a first proportion of the target population in a first number of users on the target platform, the first number of users being a number of users having at least one interaction behavior corresponding to the interaction dimension with the target account within a first preset time period. A second proportion of the target population in a second number of users on the target platform is determined, the second number of users being a number of users having at least one interaction behavior corresponding to the interaction dimension with any homogeneous account within the first preset time period. Wherein the homogeneous account is an account on the target platform having the same behavior and / or business nature as the target account, for example, the live video of the target account and the homogeneous account is a live video with goods, or the content published by both is food type, travel type or beauty type, or the product information displayed by both belongs to the same type of goods, and the like. Further, the processor 620 can determine the ratio of the first proportion to the second proportion as the single-dimension group index under the interaction dimension.

[0065] The following takes the target population as the population C, the target account as the anchor account L, and the anchor account L as an example to calculate the single-dimension group index under multiple interaction dimensions.

[0066] The single-dimension group index under the watching dimension is: the proportion of users of the population C in the number of users watching the live video of the anchor account L within the first preset time period / the proportion of users of the population C in the number of users watching the live video of any homogeneous account within the first preset time period.

[0067] The single-dimension group index under the clicking dimension is: the proportion of users of the population C in the number of users clicking the live video of the anchor account L within the first preset time period / the proportion of users of the population C in the number of users clicking the live video of any homogeneous account within the first preset time period.

[0068] The single-dimension group index under the attention dimension is: the proportion of users of the population C in the number of new fans following the anchor account L within the first preset time period / the proportion of users of the population C in the number of new fans following any homogeneous account within the first preset time period.

[0069] The single-dimension group index under the interaction dimension is: the proportion of users of the population C in the number of users producing interaction behavior with the anchor account L within the first preset time period / the proportion of users of the population C in the number of users producing interaction behavior with any homogeneous account within the first preset time period. Wherein the interaction behavior is, for example, the user inputs and sends text in the text input box of the live video.

[0070] The single-dimension group index under the participation dimension is: a proportion of users of the group C in a number of users participating in the marketing activity of the anchor account L in the first preset time period / a proportion of users of the group C in a number of users participating in the marketing activity of any homogeneous account in the first preset time period.

[0071] The calculation manner of the single-dimension group index under other dimensions is similar to the above, which is not described herein again. The first preset time period can be a time period close to the current time, for example, the last 30 days, the last 1 year, and the like.

[0072] It should be noted that each interaction dimension corresponds to one or more interaction behaviors between the target group and the target account. For example, the watching dimension corresponds to the watching interaction behavior; and under the attention dimension, the target group not only follows the target account but also watches the content published by the target account, that is, the attention dimension corresponds to the watching and following interaction behaviors; under the interaction dimension, the target group can not only interact with the target account but also follow the target account, that is, the interaction dimension corresponds to the interaction and following interaction behaviors. The above examples are illustrated by taking each interaction dimension corresponding to one interaction behavior as an example, and in some other embodiments, for some interaction dimensions with multiple interaction behaviors, the processor 620 can calculate the single-dimension group index under the interaction dimension in combination with the multiple interaction behaviors. For example, the single-dimension group index under the interaction dimension can also be: a proportion of users of the group C in a number of users following the anchor account L and interacting with the anchor account L in the first preset time period / a proportion of users of the group C in a number of users following any homogeneous account and interacting with the homogeneous account in the first preset time period.

[0073] The processor 620 can obtain a weight corresponding to each interaction dimension, and the weight reflects an interaction depth embodied by the at least one interaction behavior corresponding thereto. The deeper the interaction depth, the deeper the degree of love (preference) of the target group to the target account, and the greater the corresponding weight. The greater the weight corresponding to the interaction dimension, the greater the influence of the interaction dimension on the group tendency score; the smaller the weight corresponding to the interaction dimension, the smaller the influence of the interaction dimension on the group tendency score.

[0074] In some embodiments, the weights of the multiple interaction dimensions can be different. For example, the interaction depth of clicking is deeper than the interaction depth of watching, the interaction depth of following is deeper than the interaction depth of clicking, the interaction depth of interacting is deeper than the interaction depth of following, and the interaction depth of participating in an activity is deeper than the interaction depth of interacting, so that the weights corresponding to the watching dimension, the clicking dimension, the following dimension, the interacting dimension, and the participating-in-activity dimension have a gradually increasing trend. In some other embodiments, there are interaction dimensions corresponding to the same weight. For example, the interaction depth of interacting is the same as the interaction depth of following, so that the weights corresponding to the interacting dimension and the following dimension are the same.

[0075] The processor 620 can weight-sum each single-dimensional group index and the corresponding weight to obtain the comprehensive group index. For example, the comprehensive group index can be determined by the following formula (1):

[0076] Sum(W_i*TGI_i) (1)

[0077] wherein Sum(W_i*TGI_i) is the comprehensive group index, TGI_i is the single-dimensional group index under each interaction dimension i, and W_i is the weight corresponding to each interaction dimension i.

[0078] It should be noted that, in addition to the weighted sum method, the comprehensive group index can also be calculated by other methods, such as weighted sum and mean value. The present specification does not limit the method of calculating the comprehensive group index.

[0079] The present specification calculates the preference of the target population for the target account in the dimension, considers multiple interaction dimensions, and finally obtains a more accurate population tendency score. If the relationship between the target population and the target account is calculated based on only a single dimension, such as click rate / repurchase rate, the prediction dimension is relatively single, and the situation of the target account is not considered, resulting in inaccurate calculation results.

[0080] S144: Determine the confidence of the population tendency value, which reflects the reliability of the population tendency value.

[0081] The confidence is positively correlated with the historical interaction number of the target account in the second preset period. In some embodiments, the historical interaction number can be the number of users who interact with the target account in the second preset period. In some embodiments, the historical interaction number can be the number of users who interact with the target account and contain the target label in the tag in the second preset period. The second preset period refers to a time period before the current time, such as the last 1 year, the last half year, etc. The second preset period can be longer than the first preset period.

[0082] In some embodiments, the processor 620 can obtain a historical interaction number of the target account in a second preset time period for one or more interaction behaviors. For example, a viewing number of the target account in the past one year, an interaction number of the target account in the past half year, a number of users who both follow and interact with the target account in the past one year (i.e., each visiting user both follows and interacts), a sum of a number of users who follow and a number of users who interact with the target account in the past one year (i.e., each visiting user can follow, interact, or both follow and interact), and the like. It should be noted that the more shallow the interaction depth is, the more the historical interaction number is, for example, the historical viewing number is more than the historical following number. In order to improve the credibility of the crowd tendency score, the processor 620 can take the historical viewing number as the historical interaction number, and of course, the historical following number, the historical clicking number, and the like can also be taken as the historical interaction number, and the embodiments of the present specification are not limited thereto. The processor 620 can obtain an adjustment factor, and the confidence level can be controlled by adjusting the adjustment factor, for example, the adjustment factor is 0.01. The processor 620 can determine the confidence level based on the historical interaction number and the adjustment factor. In some embodiments, the confidence level can be calculated by the following formula (2):

[0083]

[0084] wherein arctan(P*N) is the confidence level, pi is the circular constant, arctan() is the inverse tangent function, P is the adjustment factor, and N is the historical interaction number. The value range of arctan() is Therefore, the value of the confidence level is (0, 1).

[0085] It should be noted that the confidence level can be calculated by other ways in addition to the above-mentioned way, and the embodiments of the present specification are not limited to the way of calculating the confidence level.

[0086] The embodiments of the present specification consider the influence of the historical interaction number on the crowd tendency score, the more the historical interaction number is, the more reasonable the crowd tendency score is, and the less the historical interaction number is, the less reasonable the crowd tendency score is. For example, if the historical viewing number of the target account is small, the fluctuation of the finally obtained crowd tendency score is larger, which cannot accurately reflect the degree of love of the target crowd to the target account, that is, it is less reasonable. Therefore, the historical interaction number is introduced into the calculation of the crowd tendency score in the embodiments of the present specification, the confidence level is added to the crowd tendency score through the historical interaction number, so that the calculated crowd tendency score is more credible, thereby improving the accuracy of generating the audience label for the target account.

[0087] S146: determining the crowd tendency value based on the comprehensive group index and the confidence level.

[0088] In some embodiments, the processor 620 can determine the crowd tendency value by an activation function in combination with the comprehensive crowd index and the confidence. The activation function (gate function) is, for example, a sigmoid function, a tanh function. For example, the crowd tendency value can be calculated by the following formula (3):

[0089]

[0090] wherein F is the crowd tendency value, and F() is the activation function.

[0091] The activation function can normalize the crowd tendency value, and control the value range of the crowd tendency value to [0, 1]. The closer the crowd tendency value is to 1, the more the target crowd tends to the target account, and the deeper the love degree for the target account. The closer the crowd tendency value is to 0, the less the target crowd tends to the target account, and the shallower the love degree for the target account.

[0092] In addition to determining the crowd tendency value by the method of steps S142-S146, it can also be determined by other ways, such as, only determining the crowd tendency value by the comprehensive crowd index, or further in combination with other confidence to determine the crowd tendency value, etc. The embodiments of the present application do not limit the method of determining the crowd tendency value.

[0093] After the processor 620 determines the crowd tendency value, the method P100 can further include:

[0094] S160: generating the audience label of the target account based on the target label.

[0095] If the crowd tendency value is greater than or equal to a preset threshold, the processor 620 can take the target label as the audience label of the target account, and if the crowd tendency value is less than the preset threshold, the processor 620 can not take the target label as the audience label of the target account.

[0096] All labels corresponding to the target account can be collectively referred to as understanding labels, and others can understand the target account through the understanding labels to obtain some information of the target account. The others can be access users accessing the target account, or can be back-end operation personnel. After the audience label is generated, the audience label becomes one of the understanding labels. In addition to the audience label, the understanding labels can also include personal labels related to personal information, such as gender, age, whether a star, IP address, etc., and can also include hobby labels, such as food, dressing, etc. It should be noted that the processor 620 can visually display the personal label and the hobby label near the target account, but for privacy and security considerations, the audience label can not be visually displayed, and the audience label is only used for back-end operation of the operation personnel.

[0097] In some embodiments, the processor 620 can acquire an audience label of the target account, and acquire at least one first access user who meets the audience label. The "meeting" means that the audience label is included in a plurality of labels of the account of the first access user. The first access user can be an access user who accesses the target platform before or after the audience label is generated. The processor 620 can put the target service of the target account into the account of each first access user, thereby helping the operator to achieve fine placement.

[0098] In some embodiments, the audience label can also be used to train a distribution model to achieve fine placement. For example, the processor 620 can train a distribution model (or a placement model) using a plurality of understanding labels corresponding to a plurality of registered accounts on the target platform, acquire the trained distribution model, predict a plurality of second access users who meet the understanding label of the target account using the distribution model, and then put the target service of the target account into the account of each second access user. The understanding label corresponding to each registered account includes a plurality of labels for identifying the characteristics of the registered account, including the audience label. It should be noted that the plurality of registered accounts used to train the distribution model can be homogeneous accounts of the target account, can be registered accounts with the audience label randomly selected from the target platform, or can be all registered accounts on the target platform, and the embodiments of the present specification do not limit this. It should also be noted that the processor 620 can select a few labels from the understanding labels of the registered account to train the distribution model, and the selected labels can include the audience label.

[0099] The present specification can perform single-target placement based only on the audience label, or can perform multi-target placement based on the audience label and other understanding labels, and the placement method of the target service is diversified.

[0100] To sum up, the method P100 and the system 001 for generating audience tags of a target account provided in the specification obtain a target tag from a plurality of known tags corresponding to all users of a target platform, the target tag corresponding to a target group of users in the all users, calculate a group tendency value of the target group of users with the determined tag to the target account, and then generate an audience tag of the target account based on the target tag. The target group of users in the specification is a group of users who have already possessed a determined tag. The group tendency value is determined from the perspective of the determined group tag, so that the target group of users has stronger interpretability and larger operation space. With the help of rich group tags in the target platform (such as the Alipay platform), the specification calculates the TGI value of the target account to the target group of users based on the understanding of the target account (such as the life number business anchor account), so that the generated audience tag is more accurate. The specification considers the influence of the historical interaction number (such as the historical watching number of live video) on the final group tendency value (points), adds confidence to the group tendency value bias value by introducing the historical interaction number, so that the final result is more reliable.

[0101] In another aspect of the present specification, a non-transitory storage medium storing at least one set of executable instructions for performing data processing is provided. When the executable instructions are executed by a processor, the executable instructions direct the processor to implement the steps of the method P100 for generating audience tags for target accounts described in the present specification. In some possible implementation manners, various aspects of the present specification can also be implemented in the form of a program product including program codes. When the program product is run on the system 001 for generating audience tags for target accounts, the program codes are used to cause the system 001 for generating audience tags for target accounts to perform the steps of the method P100 for generating audience tags for target accounts described in the present specification. The program product for implementing the above method can include the program codes in a portable compact disc read-only memory (CD-ROM) and can be run on the system 001. However, the program product of the present specification is not limited to this, and in the present specification, the readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system. The program product can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, be but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any appropriate combination of the above. More specific examples of the readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or a flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination of the above. The computer readable storage medium can include a data signal carried in a baseband or as part of a carrier wave propagating through a transmission medium, in which readable program codes are borne. Such a propagating data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal, or any appropriate combination of the above. The readable storage medium can also be any readable medium that is not a storage medium, which can send, propagate or transmit programs for use by or in connection with an instruction execution system, apparatus or device. The program codes contained in the readable storage medium can be transmitted in any appropriate medium, including but not limited to wireless, wired, optical fiber, RF, and the like, or any appropriate combination of the above. The program codes for performing the operations of the present specification can be written in any combination of one or more programming languages, including an object-oriented programming language, such as Java, C++, and the like, and a conventional procedural programming language, such as the "C" language or similar programming languages.The program code can execute entirely on the system 001, partly on the system 001, as a stand-alone software package, partly on the system 001 and partly on a remote computing device, or entirely on the remote computing device.

[0102] The above description of the specific embodiments of the present specification has been presented for the purpose of illustration. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still accomplish the desired results. Also, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0103] In light of the above, those skilled in the art will appreciate that the foregoing detailed description of the present specification is susceptible to various modifications and alternative forms, and that many specific details are presented for the purpose of illustration and description rather than limitation. Although the present specification has been described in detail with reference to certain implementations, variations and / or modifications can be made to the implementations described and the nomenclature used herein is meant to indicate that structures and their functionality are one example of many possible structures and their functionality. Therefore, although specific implementations have been illustrated and described herein, it should be appreciated that any arrangement can be utilized that achieves the same or similar result. This disclosure is intended to cover any and all adaptations or variations of various implementations. Therefore, it is intended that this disclosure be considered in all respects as illustrative and not restrictive, the scope of the disclosure being indicated by the appended claims rather than the foregoing description.

[0104] Furthermore, certain terminology has been used to describe embodiments of the present specification. For example, "one embodiment," "an embodiment," and / or "some embodiments” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the present specification. Therefore, it is understood that the various expressions of "in one or more embodiments," "in one embodiment,” "in an embodiment,” or "in some embodiments” are used herein to describe a particular feature, structure, or characteristic included in some embodiments of the present specification. In addition, it is understood that when a particular feature, structure, or characteristic is described in connection with any embodiment or embodiment type, it is submitted that it is within the purview of the inventor(s) to effect such feature, structure, or characteristic in connection with any other

[0105] It should be understood that in the foregoing description of embodiments of the present specification, various features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This is not necessary. Some of these features can be used to create a single integrated unit while others can be used to create more than one integrated unit. Further, not all of the features are necessarily required within a particular implementation or embodiment. Even further, an indication that a feature will be or has been "used" within certain embodiments of the present specification means that this feature can be employed within some of the embodiments of the present specification.

[0106] Each patent, patent application, publication of a patent application, and other material, for example articles, books, specifications, publications, documents, things, or the like which can be cited in the present document can be accorded with the scope of their respective copyrights. The contents of all such cited patents, patent applications, publications of patent applications, and other material are hereby incorporated by reference for all purposes to the same extent as each is accorded for its respective copyright laws. Except to the extent necessary or required to be disclaimed by law, neither the prior document history of any incorporated-by-reference material nor any prior document history of any incorporated-by- reference material is hereby incorporated by reference, and all such prior document histories are disclaimed. In the event that any of the incorporated-by-reference material contradicts any of this document, including definition or use of any term

[0107] Finally, it should be understood that the embodiments of the application disclosed herein are illustrative of the principles of the present specification. Other modifications that fall within the scope of the present specification can also be made. Thus, the embodiments disclosed in the present specification are to be considered in all respects as illustrative only and not restrictive in character. Those skilled in the art can devise numerous alternative configurations without departing from the application in the present specification. Therefore, the embodiments of the present specification are not limited to the embodiments precisely described in the application.

Claims

1. A method for generating an audience label for a target account, comprising: obtaining a target label of a target population from a plurality of population labels corresponding to a plurality of populations, wherein a total number of users of a target platform comprises access users in the plurality of populations, and each population label comprises at least one known label; determining a comprehensive population index corresponding to a plurality of interaction dimensions for the target account based on a proportion of a number of users of the target population, wherein the comprehensive population index represents a case of interaction between the target population and the target account through a target service provided by the target account in the plurality of interaction dimensions; determining a confidence of a population tendency value, wherein the confidence reflects a reliability of the population tendency value; determining the population tendency value based on the comprehensive population index and the confidence, wherein the population tendency value reflects a degree of favor of the target population to the target account; and determining the population tendency value greater than or equal to a preset threshold, and determining the target label as the audience label of the target account.

2. The method of claim 1, wherein the target service comprises at least one of the following: an online video, an offline video, an audio, an image, a text, and a commodity display. Each interaction dimension of the plurality of interaction dimensions corresponds to at least one interaction behavior between the target population and the target account.

3. The method of claim 1, wherein, 4. The method of claim 3, wherein the determining the population tendency value based on the comprehensive population index and the confidence comprises: determining the population tendency value by combining the comprehensive population index and the confidence through an activation function.

5. The method of claim 3, wherein the determining the comprehensive population index comprises: determining a single-dimensional population index of the target population interacting with the target account in each interaction dimension, respectively; obtaining a weight corresponding to each interaction dimension, wherein the weight reflects an interaction depth embodied by the at least one interaction behavior corresponding thereto; and performing a weighted summation of each single-dimensional population index and the corresponding weight to obtain the comprehensive population index.

6. The method of claim 5, wherein the determining the single-dimensional population index of the target population interacting with the target account in each interaction dimension, respectively, comprises, for each interaction dimension: determining a first proportion of a number of the target population in a first number of users on the target platform, wherein the first number of users is a number of users interacting with the target account in the at least one interaction behavior corresponding to the interaction dimension within a first preset time period; determining a second proportion of a number of the target population in a second number of users on the target platform, wherein the second number of users is a number of users interacting with any homogeneous account in the at least one interaction behavior corresponding to the interaction dimension within the first preset time period, and the homogeneous account is an account on the target platform having the same behavior and / or business nature as the target account; and determining a ratio of the first proportion to the second proportion as the single-dimensional population index in the interaction dimension. ​ ​ ​ 7. The method of claim 5, wherein, The interaction dimensions include at least one of a viewing dimension, a clicking dimension, a following dimension, an interaction dimension, and an activity participation dimension. The interaction behaviors include at least one of viewing, clicking, following, interacting, and participating in an activity.

8. The method of claim 7, wherein, The weights corresponding to the viewing dimension, the clicking dimension, the following dimension, the interaction dimension, and the activity participation dimension are increased in stages.

9. The method of claim 3, wherein, The confidence of the crowd tendency value is positively correlated with the historical interaction number of the target account in the second preset time period.

10. The method of claim 9, wherein, The confidence of the crowd tendency value is determined by: obtaining a historical interaction number of the target account in the second preset time period; obtaining an adjustment factor; and determining the confidence based on the historical interaction number and the adjustment factor.

11. The method of claim 4, wherein, The activation function is configured to normalize the crowd tendency value.

12. The method of claim 1, wherein, The target label of the target crowd is obtained from a plurality of crowd labels corresponding to a plurality of crowds by: obtaining a plurality of crowd portraits corresponding to the plurality of crowds from the target platform, each of the crowd portraits corresponding to a crowd label; and obtaining the target label from the plurality of crowd labels corresponding to the plurality of crowd portraits.

13. The method of claim 1, further comprising: obtaining the audience label of the target account; obtaining a plurality of first access users meeting the audience label; and delivering the target service of the target account to an account of each of the first access users.

14. The method of claim 1, further comprising: obtaining a distribution model trained by a plurality of understanding labels of a plurality of registered accounts of the target platform, each of the understanding labels including a plurality of labels for identifying characteristics of the registered account, and each of the understanding labels including the audience label; predicting a plurality of second access users meeting the understanding label of the target account using the distribution model; and delivering the target service of the target account to an account of each of the second access users.

15. A system for generating an audience label for a target account, comprising: at least one storage medium storing at least one set of instructions for generating an audience label for a target account; and at least one processor in communication connection with the at least one storage medium, wherein when the system for generating an audience label for a target account is running, the at least one processor reads the at least one set of instructions and implements the method for generating an audience label for a target account according to any one of claims 1-14. ​ ​

Images