Information push method, device, server and computer-readable storage medium

Abstract

The present application provides an information push method, device, server and computer-readable storage medium, which relate to the field of Internet technology, wherein the method comprises: determining an estimated click-through rate of the recommended information of the recommended topic based on characteristic information of the recommended topic and characteristic information of global users; determining a target user from the global users based on the estimated click-through rate; and sending the recommended information of the recommended topic to the target user. The embodiment of the present application selects target users with a higher probability of clicking on the recommended message from global users based on the estimated click-through rate, and can selectively send recommended messages to different users, thereby improving the click-through rate and conversion rate of the recommended message.

Description

Technical Field

[0001] The present application relates to the field of Internet technology, and more specifically, to an information push method, device, server, and computer-readable storage medium. Background Art

[0002] With the continuous development of the Internet, the traffic distribution platform can send recommendation information to each user in the platform, so that the user can read the received recommendation information.

[0003] In the related art, a large amount of recommended information can be pushed to multiple users of the platform to improve the click-through rate and conversion rate of each recommended information. However, in the case of limited traffic resources, the recommended information can only be promoted to a limited number of users.

[0004] Therefore, there is an urgent need for an information push method that can improve the click-through rate and conversion rate of each recommended information under the condition of limited traffic resources. Summary of the invention

[0005] In view of this, the purpose of the embodiments of the present application is to provide an information push method, device, server and computer-readable storage medium, which can determine the estimated click-through rate through local users, thereby selecting target users through the estimated click-through rate, and improving the click-through rate and conversion rate of recommended information.

[0006] According to a first aspect of the present application, there is provided an information push method, comprising:

[0007] Determining an estimated click rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the global user;

[0008] Determining a target user from the global users according to the estimated click rate;

[0009] Sending recommendation information of the recommended topic to the target user.

[0010] In some embodiments, determining the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user includes:

[0011] Randomly select multiple users from the global users as local users;

[0012] According to the characteristic information of the recommended topic and the characteristic information of the local users, an observation parameter of the local users is determined by a trajectory tracking method, where the observation parameter is used to represent the traffic resources corresponding to each user in the local users;

[0013] An estimated click rate of the recommended information of the recommended topic is determined according to the feature information of the recommended topic and the observation parameter of the local user.

[0014] In some embodiments, determining the observation parameters of the local user by means of trajectory tracking according to the feature information of the recommended topic and the feature information of the local user includes:

[0015] According to the feature information of the recommended topic and the feature information of the local users, a feature distance between the feature information of the recommended topic and the feature information of each of the local users, and a feature change speed between the feature information of the recommended topic and the feature information of each of the local users are determined by using a trajectory tracking method;

[0016] An observation parameter of the local user is determined according to the characteristic distance and the characteristic change speed.

[0017] In some embodiments, determining the estimated click rate of the recommended information of the recommended topic according to the feature information of the recommended topic and the observation parameter of the local user includes:

[0018] Determining whether the observation parameters of the local user meet preset constraints;

[0019] If the observation parameter of the local user does not satisfy the constraint condition, updating the observation parameter of the local user according to a preset step size control parameter, a preset weight of the feature change speed, the feature distance and the feature change speed;

[0020] An estimated click rate of the recommended information of the recommended topic is determined according to the characteristic information of the recommended topic and the updated observation parameter of the local user.

[0021] In some embodiments, determining the target user from the global users according to the estimated click rate includes:

[0022] Determining an estimated engagement level of each user in the global users according to the estimated click rate and a preset engagement level;

[0023] According to the estimated participation degree of each user in the global users, at least one user with the highest estimated participation degree is selected from the global users as the target user.

[0024] In some embodiments, before determining the estimated click rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the global user, the method further includes:

[0025] Acquiring initial feature information of the recommended topic;

[0026] The initial feature information is mapped to obtain feature information of the recommended topic, wherein the dimension of the feature information of the recommended topic is greater than the dimension of the initial feature information.

[0027] In some embodiments, before mapping the initial feature information to obtain the feature information of the recommended topic, the method further includes:

[0028] Normalizing the initial feature information to obtain normalized initial feature information;

[0029] The mapping of the initial characteristic information to obtain the characteristic information of the recommended topic includes:

[0030] The normalized initial feature information is mapped to obtain feature information of the recommended topic.

[0031] In some embodiments, the characteristic information of the recommended topic includes at least one of the following types of characteristic information: user portrait information, order characteristic information, access characteristic information, activity voucher attribute information, document characteristic information, and historical topic characteristic information.

[0032] According to a second aspect of the present application, there is provided an information push device, comprising:

[0033] A click rate determination module, used to determine the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user;

[0034] A user determination module, used to determine a target user from the global users according to the estimated click rate;

[0035] The sending module is used to send the recommendation information of the recommended topic to the target user.

[0036] In some embodiments, the click-through rate determination module is also used to randomly select multiple users from the global users as local users; determine the observation parameters of the local users by trajectory tracking based on the characteristic information of the recommended topic and the characteristic information of the local users, and the observation parameters are used to represent the traffic resources corresponding to each user in the local users; determine the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the observation parameters of the local users.

[0037] In some embodiments, the click-through rate determination module is also used to determine the characteristic distance between the characteristic information of the recommended topic and the characteristic information of each local user, as well as the characteristic change speed between the characteristic information of the recommended topic and the characteristic information of each local user by trajectory tracking based on the characteristic information of the recommended topic and the characteristic information of the local user; and determine the observation parameters of the local user based on the characteristic distance and the characteristic change speed.

[0038] In some embodiments, the click-through rate determination module is also used to determine whether the observation parameters of the local user meet preset constraints; if the observation parameters of the local user do not meet the constraints, the observation parameters of the local user are updated according to preset step control parameters, preset weights of feature change speed, the feature distance and the feature change speed; based on the feature information of the recommended topic and the updated observation parameters of the local user, the estimated click-through rate of the recommended information of the recommended topic is determined.

[0039] In some embodiments, the user determination module is further used to determine the estimated engagement of each user among the global users based on the estimated click-through rate and the pre-set preset engagement; and based on the estimated engagement of each user among the global users, select at least one user with the highest estimated engagement from the global users as the target user.

[0040] In some embodiments, the apparatus further comprises:

[0041] An acquisition module, used to acquire initial feature information of the recommended topic;

[0042] A mapping module is used to map the initial feature information to obtain feature information of the recommended topic, wherein the dimension of the feature information of the recommended topic is greater than the dimension of the initial feature information.

[0043] In some embodiments, the apparatus further comprises:

[0044] A normalization module, used for normalizing the initial feature information to obtain normalized initial feature information;

[0045] The mapping module is further used to map the normalized initial feature information to obtain the feature information of the recommended topic.

[0046] In some embodiments, the characteristic information of the recommended topic includes at least one of the following types of characteristic information: user portrait information, order characteristic information, access characteristic information, activity voucher attribute information, document characteristic information, and historical topic characteristic information.

[0047] According to the third aspect of the present application, a server is provided, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the server is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the information push method as described in any one of the first aspects.

[0048] According to a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the information push method as described in any one of the first aspects are executed.

[0049] Based on any of the above aspects, in the information push method, device, server and computer-readable storage medium provided by the present application, the estimated click rate of the recommended information of the recommended topic is determined based on the characteristic information of the recommended topic and the characteristic information of the global users, and the target user is determined from the global users based on the estimated click rate, and then the recommended information of the recommended topic is sent to the target user. It is achieved that the target user with a high probability of clicking the recommended message is selected from the global users based on the estimated click rate, and the recommended message can be selectively sent to different users, thereby improving the click rate and conversion rate of the recommended message. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A block diagram of an information push system provided by an embodiment of the present application is shown;

[0052] Figure 2 A schematic diagram showing a flow chart of an information push method provided by an embodiment of the present application;

[0053] Figure 3 A schematic diagram showing a flow chart of an information push method provided by another embodiment of the present application;

[0054] Figure 4 A schematic diagram showing a flow chart of an information push method provided by another embodiment of the present application;

[0055] Figure 5 A block diagram of an information push device provided by an embodiment of the present application is shown;

[0056] Figure 6A block diagram of an information push device provided by another embodiment of the present application is shown;

[0057] Figure 7 A block diagram of an information push device provided by another embodiment of the present application is shown;

[0058] Figure 8 A schematic diagram of the structure of a server provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0059] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0060] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0061] In order to enable those skilled in the art to use the contents of this application, the following implementation methods are provided in conjunction with the specific application scenario "information push". For those skilled in the art, the general principles defined herein can be applied to other embodiments and application scenarios without departing from the spirit and scope of this application. Although this application is mainly described around information push, it should be understood that this is only an exemplary embodiment. For example, information push can be applied to various scenarios on the Internet, such as a client pushing information to a user, a web page pushing information to a client, and other methods may also be included, such as pushing various types of information to users via text messages, instant messaging, etc., which are not limited by this application.

[0062] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0063] Before the present application was filed, different recommended information could not be accurately pushed to different users under limited traffic resources, resulting in low click-through rate and conversion rate of recommended information. The present application provides an information push method to achieve targeted push of corresponding push information to target users.

[0064] Figure 1 A block diagram of an information push system provided by an embodiment of the present application is shown. For example, the information push system 100 can be applied to information push platforms such as taxis, chauffeur services, express trains, carpooling, bus services, driver rentals, or shuttle services. The information push system 100 may include one or more of a server 110, a network 120, a service requester terminal 130, a service provider terminal 140, and a database 150, and the server 110 may include a processor for executing instruction operations. Of course, in some application scenarios, the service requester terminal 130 and the service provider terminal 140 may not be distinguished, and the user terminal may be used as a general term.

[0065] In some embodiments, the server 110 may be a single server or a server group. The server group may be centralized or distributed (for example, the server 110 may be a distributed system). In some embodiments, the server 110 may be local or remote relative to the terminal. For example, the server 110 may access information and / or data stored in the service requester terminal 130, the service provider terminal 140, or the database 150, or any combination thereof via the network 120. As another example, the server 110 may be directly connected to at least one of the service requester terminal 130, the service provider terminal 140, and the database 150 to access the stored information and / or data. In some embodiments, the server 110 may be implemented on a cloud platform; as an example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an inter-cloud, a multi-cloud, etc., or any combination thereof.

[0066] In some embodiments, the server 110 may include a processor. The processor may process information and / or data related to information push to perform one or more functions described in the present application. For example, the processor may select users with a higher probability of clicking on recommended information from a large number of users as target users based on the data stored in the database 150. In some embodiments, the processor may include one or more processing cores (e.g., a single-core processor (S) or a multi-core processor (S)). By way of example only, the processor may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), or a microprocessor, or any combination thereof.

[0067] The network 120 may be used for the exchange of information and / or data. In some embodiments, one or more components in the information push system 100 (e.g., the server 110, the service requester terminal 130, the service provider terminal 140, and the database 150) may send information and / or data to other components. For example, the server 110 may send recommendation information to the service requester terminal 130 or the service provider terminal 140 corresponding to the target user via the network 120. In some embodiments, the network 120 may be any type of wired or wireless network, or a combination thereof. As just an example, the network 120 may include a wired network, a wireless network, a fiber optic network, a telecommunications 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, or a near field communication (NFC) network, or any combination thereof. In some embodiments, the network 120 may include one or more network access points. For example, the network 120 may include a wired or wireless network access point, such as a base station and / or a network switching node, through which one or more components of the information push system 100 may connect to the network 120 to exchange data and / or information.

[0068] In some embodiments, the service requester terminal 130 may include a mobile device, a tablet computer, a laptop computer, or a built-in device in a motor vehicle, etc., or any combination thereof.

[0069] In some embodiments, the service provider terminal 140 may be a device similar or identical to the service requester terminal 130. In some embodiments, the service provider terminal 140 may be a device with positioning technology for locating the location of the service provider and / or the service provider terminal. In some embodiments, the service requester terminal 130 and / or the service provider terminal 140 may communicate with other positioning devices to determine the location of the service requester, the service requester terminal 130, the service provider, or the service provider terminal 140, or any combination thereof. In some embodiments, the service requester terminal 130 and / or the service provider terminal 140 may send positioning information to the server 110.

[0070] The database 150 may store data and / or instructions. In some embodiments, the database 150 may store data obtained from the service requester terminal 130 and / or the service provider terminal 140. For example, the database 150 may store user information (such as user information and identification information, etc.) registered by the user through the service requester terminal 130 and / or the service provider terminal 140 and theme activities participated by the user (such as short-term vehicle rental theme activities and refueling day theme activities, etc.). In some embodiments, the database 150 may store data and / or instructions of the exemplary methods described in this application. In some embodiments, the database 150 may include a large-capacity memory, a removable memory, a volatile read-write memory, or a read-only memory (ROM), etc., or any combination thereof. As an example, mass storage may include magnetic disks, optical disks, solid-state drives, etc.; removable storage may include flash drives, floppy disks, optical disks, memory cards, zip disks, magnetic tapes, etc.; volatile read-write storage may include random access memory (Random Access Memory, RAM); RAM may include dynamic RAM (Dynamic Random Access Memory, DRAM), double data rate synchronous dynamic RAM (Double Date-Rate Synchronous RAM, DDR SDRAM); static RAM (Static Random-Access Memory, SRAM), thyristor RAM (Thyristor-Based Random Access Memory, T-RAM) and zero capacitor RAM (Zero-RAM), etc. As an example, ROM may include mask ROM (Mask Read-Only Memory, MROM), programmable ROM (Programmable Read-Only Memory, PROM), erasable programmable ROM (Programmable Erasable Read-only Memory, PEROM), electrically erasable programmable ROM (Electrically Erasable Programmable read only memory, EEPROM), CD-ROM (CD-ROM), and digital versatile disk ROM, etc. In some embodiments, database 150 may be implemented on a cloud platform. As an example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, a cross-cloud, a multi-cloud, or other similar, etc., or any combination thereof.

[0071] In some embodiments, the database 150 may be connected to the network 120 to communicate with one or more components in the information push system 100 (e.g., the server 110, the service requester terminal 130, the service provider terminal 140, etc.). One or more components in the information push system 100 may access data or instructions stored in the database 150 via the network 120. In some embodiments, the database 150 may be directly connected to one or more components in the information push system 100 (e.g., the server 110, the service requester terminal 130, the service provider terminal 140, etc.); or, in some embodiments, the database 150 may also be part of the server 110.

[0072] In some embodiments, one or more components in the information push system 100 (e.g., server 110, service requester terminal 130, service provider terminal 140, etc.) may have access to the database 150. In some embodiments, when certain conditions are met, one or more components in the information push system 100 may read and / or modify information related to the service requester, the service provider, or the public, or any combination thereof. For example, the server 110 may read and / or modify information of one or more users after receiving a service request.

[0073] It should be noted that, during the execution of the information push method described in the embodiment of the present application, the server 110 may send recommendation information corresponding to the recommended topic to the terminal corresponding to the target user according to the identification information of the determined target user.

[0074] Figure 2 The flowchart of the information push method provided by an embodiment of the present application is shown in FIG. Figure 1 The server shown in . Figure 2 As shown, the method includes:

[0075] Step 201: Determine an estimated click rate of recommended information of the recommended topic based on characteristic information of the recommended topic and characteristic information of global users.

[0076] Global users may refer to all users registered on a certain platform or multiple platforms, and target users to whom push information corresponding to a recommended topic is selected from among these users.

[0077] The recommended information may be information pushed to the user to introduce related information such as theme activities, consultations, services, etc. corresponding to the recommended theme. The estimated click rate may indicate the probability that each user may click to view the recommended information.

[0078] In an optional implementation, the server may first obtain feature information of the recommended topic, then obtain feature information of each global user, and then compare and analyze the user's feature information with the feature information of the recommended topic to obtain an estimated click rate of each user's recommended information for the recommended topic.

[0079] Step 202: Determine target users from global users based on the estimated click rate.

[0080] After obtaining the estimated click-through rate, the effect of users' participation in the theme activities corresponding to the recommended topics after sending recommendation messages to each user can be estimated based on the estimated click-through rate of each user. Then, in the case of limited traffic resources, a preset number of users can be selected as target users based on the effect of each user's participation in the theme activities corresponding to the recommended topics, so that in subsequent steps, recommendation information can be sent to the target users.

[0081] Step 203: Send recommendation information of the recommended topic to the target user.

[0082] After selecting the target users, the server can send recommendation information to each target user according to the identification information of each target user, so that the user who receives the recommendation information clicks on the recommendation information to view it, thereby improving the click-through rate and conversion rate of the recommendation information.

[0083] Among them, the user's identification information can be ID (Identification) information (such as user name), contact information (such as mobile phone number, third-party application number), etc. when registering on the platform, and this embodiment of the application does not limit this.

[0084] It should be noted that, in actual applications, recommendation information can be sent to users based on multiple recommendation topics, but this application only takes one recommendation topic as an example for explanation, and does not limit the number of recommendation topics.

[0085] For example, recommendation information corresponding to a certain recommendation topic may be sent to multiple users, but recommendation information corresponding to another recommendation topic may not be sent to any user due to traffic resource limitations.

[0086] In summary, the information push method provided in the embodiment of the present application determines the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the local users in the global users, determines the target user from the global users based on the estimated click-through rate, and then sends the recommended information of the recommended topic to the target user. It is possible to selectively send more matching recommended messages to different users, thereby improving the click-through rate and conversion rate of the recommended messages.

[0087] Figure 3 A flow chart of an information push method provided by another embodiment of the present application is shown. On the basis of the above embodiment, in order to reduce the amount of calculation, local users can be further screened from global users for calculation. In an optional implementation, the server can search for users in the global users whose characteristic information is similar to that of each user in the local users based on the characteristic information and estimated click-through rate of each user in the local users, thereby taking the estimated click-through rate of a certain user in the local users as the estimated click-through rate of users similar to the user, and then estimating the estimated click-through rate of each user in the global users. For example, for each user in the local users, the server can first obtain the characteristic information of the user, and search for users in the global users who match the characteristic information of the user. If a user with similar characteristic information to the user is found in the global users, the estimated click-through rate of the local user can also be used as the estimated click-through rate of similar users in the global users.

[0088] like Figure 3 As shown, the above-mentioned determination of the estimated click rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the global user includes:

[0089] Step 301: randomly select multiple users from global users as local users.

[0090] Since the calculation is performed using the characteristic information of global users, the amount of calculation is huge. In order to save computing resources and avoid long calculation time, some users can be selected from the global users as local users, and the characteristic information of the recommended topic and the characteristic information of each user in the local users are matched to obtain the estimated click rate of the recommended information for the recommended topic.

[0091] For example, in the process of randomly selecting local users, one hundred thousand users may be randomly selected from tens of millions of global users as local users.

[0092] Step 302: According to the feature information of the recommended topic and the feature information of the local user, the observation parameters of the local user are determined by using a trajectory tracking method.

[0093] The observation parameter is used to represent the traffic resources corresponding to each user in the local user group. For example, the traffic of the recommended topic and the accuracy of the estimated click rate can be adjusted by adjusting the size of the observation parameter. If the observation parameter is large, the recommended topic can obtain a large amount of traffic, but the accuracy of the estimated click rate is low.

[0094] In an optional implementation, the server may first randomly select local users from global users, and compare and analyze the feature information of each local user with the feature information of the recommended topic to obtain the observation parameters of the local users.

[0095] Furthermore, in the process of determining the observation parameters, the characteristic distance between the characteristic information of the recommended topic and the characteristic information of each local user, as well as the characteristic change rate between the characteristic information of the recommended topic and the characteristic information of each local user can be determined by trajectory tracking based on the characteristic information of the recommended topic and the characteristic information of the local user, and then the observation parameters of the local user can be determined based on the characteristic distance and the characteristic change rate.

[0096] For example, the feature information of the recommended topic and the feature information of each user may be compared and calculated to obtain the Euclidean distance between the two feature information, and the Euclidean distance may be used as the feature distance between the feature information of the recommended topic and the feature information of each user in the local user group.

[0097] Step 303: Determine the estimated click rate of the recommended information of the recommended topic based on the feature information of the recommended topic and the observation parameters of the local user.

[0098] After obtaining the observation parameters of the local users, the estimated click rate of the recommended information of the recommended topic can be analyzed and calculated based on the observation parameters and the characteristic information of the recommended topic. However, the observation parameters need to meet certain constraints. Therefore, before determining the estimated click rate, it can be determined whether the observation parameters meet the preset constraints.

[0099] If the observed parameters do not satisfy the preset constraints, the observed parameters need to be updated until the updated observed parameters satisfy the constraints, and then the estimated click rate can be calculated based on the updated observed parameters.

[0100] Optionally, it is possible to first determine whether the observation parameters of the local user satisfy preset constraints. If the observation parameters of the local user do not satisfy the constraints, the observation parameters of the local user may be updated according to preset step control parameters, preset weights of the feature change rate, the feature distance, and the feature change rate, and then determine the estimated click rate of the recommended information of the recommended topic based on the feature information of the recommended topic and the updated observation parameters of the local user.

[0101] For example, the formula w ij =1 and λ≥0 to determine the observation parameter λ k Whether the constraints are met, Used to indicate that when k is equal to j, and The maximum value can be obtained, where C ik represents the click-through rate of the recommended information of the kth recommended topic by the i-th user, λ k represents the observation parameter corresponding to the kth recommended topic, Rk represents the preset participation level corresponding to the kth recommended topic, L is the number of deterministic recommended topics, and M is the number of recommended information to be sent.

[0102] The preset participation level is used to indicate the possibility of a user participating in the activity corresponding to the recommended topic after receiving the recommendation information. Optionally, the preset benefits corresponding to the recommended topic can be evaluated based on the participation level, for example, R k It can represent the preset benefit corresponding to the k-th recommended topic, which is not specifically limited here.

[0103] It should be noted that deterministic recommendation topics can represent recommended topics for activities with a relatively high priority and a high probability of being executed. For example, the Double Eleven event must have 2,000 participants, so we need to ensure that there are at least 2,000 users in the event. Some activities are actively participated by users, so we also need to ensure that the activities that users have signed up for are held.

[0104] If the observed parameters do not satisfy the constraints, then the formula λ j1 =λ j0 -α*(ΔS j +βΔv j ) updates the observation parameters to obtain the updated observation parameters. j1 is the updated observation parameter, λ j0 is the observation parameter before updating, ΔS j is the characteristic distance, Δv j is the characteristic change speed, α is the preset step length control parameter, and β is the preset weight of the characteristic change speed. For example, α may be 0.2, and β may be 0.8.

[0105] In summary, the information push method provided in the embodiment of the present application randomly selects multiple users from global users as local users, and determines the observation parameters of the local users by using trajectory tracking according to the characteristic information of the recommended topic and the characteristic information of the local users, and finally determines the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the observation parameters of the local users. By determining the estimated click rate for the local users among the global users, the problem of being unable to optimize for a large amount of data can be avoided, the amount of computing power of the server is reduced, and the efficiency of determining the estimated click rate is improved.

[0106] Figure 4 FIG. 1 is a flow chart showing a method for pushing information provided by another embodiment of the present application. The execution subject of the method may be as follows: Figure 1 The server shown in . Figure 4 As shown, the method includes:

[0107] Step 401: Obtain feature information of recommended topics.

[0108] The characteristic information of the recommended topic may include at least one of the following types of characteristic information: user portrait information, order characteristic information, access characteristic information, activity voucher attribute information, document characteristic information and historical topic characteristic information.

[0109] In order to select target users with a high probability of clicking on the recommended information from global users, a comparative analysis can be performed based on the feature information of each user and the feature information of the recommended topic to determine the target users who are interested in the recommended information and have a high probability of clicking on the recommended information.

[0110] Before performing comparative analysis on the characteristic information, the server may first obtain characteristic information for the recommended topic, so that in subsequent steps, the target user may be determined according to the characteristic information of the recommended topic.

[0111] The information platform for sending recommendation information to users may store data of multiple recommendation topics. If recommendation information of a certain recommendation topic needs to be pushed, the data of the recommendation topic may be obtained first, and then characteristic information of the recommendation topic may be extracted based on the data of the recommendation topic.

[0112] Optionally, the server may obtain initial feature information of the recommended topic, and map the initial feature information to obtain feature information of the recommended topic, wherein the dimension of the feature information of the recommended topic is greater than the dimension of the initial feature information.

[0113] In an optional implementation, the server can extract and quantify the data of recommended topics in the information platform to obtain initial feature information, and then map the initial feature information according to the mapping method of the neural network to obtain feature information of feature topics with higher dimensions and richer information.

[0114] Furthermore, in order to reduce the amount of computation on the server, the initial feature information can be normalized before being mapped to obtain normalized initial feature information. Correspondingly, in the process of mapping the feature information, the normalized initial feature information can be mapped to obtain feature information of the recommended topic.

[0115] Step 402: Determine the estimated click rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the global user.

[0116] Optionally, after obtaining the characteristic information of the recommended topic, local users may be randomly selected from the global users, and the estimated click rate of the recommended information for the recommended topic may be determined based on the selected local users.

[0117] Since the process of determining the estimated click rate in step 402 is similar to the process of determining the estimated click rate in steps 301 to 303, it will not be repeated here.

[0118] Step 403: Determine the target user from the global users according to the estimated click rate.

[0119] After obtaining the estimated click-through rate, the estimated engagement of each user after sending the recommendation information to each user can be calculated based on the estimated click-through rate and the pre-set preset engagement and estimated engagement algorithm. Therefore, based on the size of each estimated engagement, the users corresponding to the maximum estimated engagement can be selected as target users in limited traffic resources.

[0120] The estimated participation degree may include the possibility of the user actually participating in the activity, the possibility of consuming or participating in related activities, etc. That is, the estimated participation degree is used to indicate the possibility of the user participating in the theme activity of the recommended theme corresponding to the recommendation information after receiving the recommendation information. For example, if a user receives the recommendation information, clicks on the recommendation information, and will definitely participate in the theme activity corresponding to the recommendation information, then the estimated participation degree corresponding to the user may be 1. Similarly, if the user will definitely not participate in the theme activity corresponding to the recommendation information, then the estimated participation degree corresponding to the user may be 0.

[0121] Optionally, the server may determine the estimated engagement of each user in the global users based on the estimated click rate and a preset engagement, and then select at least one user with the highest estimated engagement from the global users as the target user based on the estimated engagement of each user in the global users.

[0122] In an optional implementation, the estimated click rate of each user can be obtained, and combined with a pre-designed preset engagement, calculated using a pre-set estimated engagement algorithm, and the parameter values ​​corresponding to the estimated click rate and the preset engagement are respectively substituted into the estimated engagement algorithm to obtain the estimated engagement of each user, and then based on limited traffic resources, at least one user corresponding to the maximum estimated engagement is selected as the target user.

[0123] Furthermore, in actual applications, target users can be selected based on multiple recommended topics. However, in order to avoid disturbing users, only one recommendation message can be sent to one user, thereby limiting the method of determining target users. In addition, for some recommended topics with traffic requirements, the method of determining target users can also be limited based on the traffic requirements of the recommended topics.

[0124] Among them, the recommended topics can correspond to multiple aspects. For example, the recommended topics can correspond to multiple aspects including car rental, time-sharing, refueling, maintenance, charging and car life, and each aspect of the content can include multiple recommended topics. For example, the recommended topics corresponding to the car rental content can include financial leasing theme activities and short-term rental theme activities; the recommended topics corresponding to the time-sharing content can include red envelope theme activities and combination theme activities; the recommended topics corresponding to the refueling content can include learning theme activities and refueling day theme activities; the recommended topics corresponding to the maintenance content can include telemarketing theme activities and automatic marketing theme activities; the recommended topics corresponding to the charging content can include special price lottery theme activities and Kaicheng letter theme activities; the recommended topics corresponding to the car life content can include community operation theme activities.

[0125] Corresponding to the above process of selecting at least one user corresponding to the maximum estimated engagement as the target user, based on the recommendation information of multiple recommended topics, an allocation matrix can be generated according to each recommended topic and each user, and the estimated engagement generated by each user clicking on the recommendation information of each recommended topic can be calculated through the estimated engagement algorithm, so that the allocation matrix can be filled according to each estimated engagement, so that a target user can be selected from at least one user in limited traffic resources, so that the estimated engagement of sending recommendation information to the target user is maximized.

[0126] For example, if the number of global users is 50 million and the number of recommended topics is 100, a distribution matrix W can be generated for global users and multiple recommended topics. The row vectors (or column vectors) of the distribution matrix W can be users, and the column vectors (or row vectors) can be recommended topics. The distribution matrix W is filled by the estimated engagement generated by whether each user clicks on each recommended topic. Combined with limited traffic resources, the above-mentioned limitation of sending only one recommendation message to one user and recommended topics with traffic requirements, the distribution matrix W is formalized as an integer programming problem, and finally the target users among global users are obtained by solving it.

[0127] Optionally, the participation algorithm can be estimated by the formula corresponding to Solve to get the above target users, and the constraints can include ∑ j w ij -1=0 and b j ε-∑ i w ij ≤0, where w ij It is used to indicate whether the i-th user clicks on the recommended information of the j-th recommended topic. If the user clicks on w ij 1, do not click w ij is 0; C ij It is used to represent the probability that the i-th user clicks on the recommended information of the j-th recommended topic, Rj It is used to indicate the preset engagement degree corresponding to clicking the recommendation information of the jth recommended topic, b j is the traffic requirement of the jth recommended topic, ε is the traffic satisfaction rate, N is the number of global users, and M is the limited traffic resource, for example, the total amount of recommended messages that can be sent. In summary, the formula can be determined It is used to represent the minimum value of the estimated participation with negative values, that is, the maximum value of the estimated participation with positive values. By determining the maximum value of the estimated participation, w can be determined. ij Whether the i-th user needs to click on the recommended information of the j-th recommended topic, so as to ij The i-th user whose value is 1 is determined as the target user.

[0128] Step 404: Send recommendation information of the recommended topic to the target user.

[0129] It should be noted that in actual applications, through AB experiments (a kind of experimental method), it can be found that after sending recommendation information to target users using the above method, the overall click rate of recommendation information has been increased by 20%. Moreover, using the above method, it only takes 2 hours to solve the target user in the decision variables of the order of tens of billions.

[0130] In summary, the information push method provided in the embodiment of the present application determines the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the local users in the global users, determines the target user from the global users based on the estimated click-through rate, and then sends the recommended information of the recommended topic to the target user. By selecting the target user with a high probability of clicking the recommended message from the global users based on the estimated click-through rate, the recommended message can be selectively sent to different users, thereby improving the click-through rate and conversion rate of the recommended message.

[0131] Furthermore, by determining the estimated click rate for local users among the global users, the problem of being unable to optimize for a large amount of data can be avoided, the amount of computing on the server is reduced, and the efficiency of determining the estimated click rate is improved.

[0132] Figure 5 A block diagram of an information push device provided by an embodiment of the present application is shown, and the functions implemented by the information push device correspond to the steps performed by the above method. The device can be understood as the above server, or a processor of the server, or can be understood as a component independent of the above server or processor that implements the functions of the present application under the control of the server. As shown in the figure, the information push device can include: a click rate determination module 501, a user determination module 502, and a sending module 503.

[0133] The click rate determination module 501 is used to determine the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user;

[0134] The user determination module 502 is used to determine a target user from the global users according to the estimated click rate;

[0135] The sending module 503 is used to send the recommendation information of the recommended topic to the target user.

[0136] Optionally, the click-through rate determination module 501 is also used to randomly select multiple users from the global users as local users; determine the observation parameters of the local users by trajectory tracking based on the characteristic information of the recommended topic and the characteristic information of the local users, and the observation parameters are used to represent the traffic resources corresponding to each user in the local users; determine the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the observation parameters of the local users.

[0137] Optionally, the click-through rate determination module 501 is also used to determine the characteristic distance between the characteristic information of the recommended topic and the characteristic information of each user in the local users, and the characteristic change speed between the characteristic information of the recommended topic and the characteristic information of each user in the local users by trajectory tracking based on the characteristic information of the recommended topic and the characteristic information of the local users; and determine the observation parameters of the local user based on the characteristic distance and the characteristic change speed.

[0138] Optionally, the click-through rate determination module 501 is also used to determine whether the observation parameters of the local user meet preset constraints; if the observation parameters of the local user do not meet the constraints, the observation parameters of the local user are updated according to the preset parameter control step size, the preset weight of the feature change speed, the feature distance and the feature change speed; based on the feature information of the recommended topic and the updated observation parameters of the local user, the estimated click-through rate of the recommended information of the recommended topic is determined.

[0139] Optionally, the user determination module 502 is further used to determine the estimated engagement of each user in the global user group based on the estimated click-through rate and a pre-set preset engagement; and based on the estimated engagement of each user in the global user group, select at least one user with the highest estimated engagement from the global user group as the target user.

[0140] Alternatively, see Figure 6 , the device further comprises:

[0141] An acquisition module 504 is used to acquire initial feature information of the recommended topic;

[0142] The mapping module 505 is used to map the initial feature information to obtain feature information of the recommended topic, where the dimension of the feature information of the recommended topic is greater than the dimension of the initial feature information.

[0143] Alternatively, see Figure 7 , the device further comprises:

[0144] A normalization module 506 is used to normalize the initial feature information to obtain normalized initial feature information;

[0145] The mapping module 505 is further used to map the normalized initial feature information to obtain feature information of the recommended topic.

[0146] Optionally, the feature information of the recommended topic includes at least one of the following types of feature information: user portrait information, order feature information, access feature information, activity voucher attribute information, document feature information and historical topic feature information.

[0147] In summary, the information push device provided in the embodiment of the present application determines the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the local users in the global users, determines the target user from the global users based on the estimated click-through rate, and then sends the recommended information of the recommended topic to the target user. By selecting the target user with a high probability of clicking the recommended message from the global users based on the estimated click-through rate, the recommended message can be selectively sent to different users, thereby improving the click-through rate and conversion rate of the recommended message.

[0148] The above modules can be connected or communicated with each other via wired connection or wireless connection. The wired connection may include metal cables, optical cables, hybrid cables, etc., or any combination thereof. The wireless connection may include a connection in the form of LAN, WAN, Bluetooth, ZigBee, or NFC, or any combination thereof. Two or more modules can be combined into a single module, and any module can be divided into two or more units. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the method embodiment, and will not be repeated in this application. In several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the coupling or direct coupling or communication connection between each other shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0149] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0150] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0151] It should be noted that the above modules can be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital singnal processors (DSP), or one or more field programmable gate arrays (FPGA). For another example, when a module above is implemented in the form of a processing element scheduling program code, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processors that can call program codes. For another example, these modules can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0152] Figure 8 A schematic diagram of the structure of a server provided in an embodiment of the present application is as follows: Figure 8 As shown, the device includes: a processor 801, a storage medium 802 and a bus 803, wherein:

[0153] The storage medium 802 stores machine-readable instructions executable by the processor 801. When the server is running, the processor 801 communicates with the storage medium 802 via the bus 803, and the processor 801 executes the machine-readable instructions to perform the above method embodiment. The specific implementation method and technical effect are similar and will not be repeated here.

[0154] Optionally, the present invention further provides a program product, such as a computer-readable storage medium, on which a computer program is stored, and the computer program is used to execute the above method embodiment when executed by a processor.

[0155] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An information push method, characterized in that: include: Determining an estimated click rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the characteristic information of the global user; Determining a target user from the global users according to the estimated click rate; Sending recommendation information of the recommended topic to the target user; The step of determining the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user includes: Randomly select multiple users from the global users as local users; According to the characteristic information of the recommended topic and the characteristic information of the local users, an observation parameter of the local users is determined by a trajectory tracking method, where the observation parameter is used to represent the traffic resources corresponding to each user in the local users; Determining an estimated click rate of the recommended information of the recommended topic according to the feature information of the recommended topic and the observation parameter of the local user; The step of determining the observation parameters of the local user by using a trajectory tracking method according to the feature information of the recommended topic and the feature information of the local user includes: According to the feature information of the recommended topic and the feature information of the local users, a feature distance between the feature information of the recommended topic and the feature information of each of the local users, and a feature change speed between the feature information of the recommended topic and the feature information of each of the local users are determined by using a trajectory tracking method; An observation parameter of the local user is determined according to the characteristic distance and the characteristic change speed.

2. The method according to claim 1, characterized in that The step of determining the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the observation parameter of the local user includes: Determining whether the observation parameters of the local user meet preset constraints; If the observation parameter of the local user does not satisfy the constraint condition, updating the observation parameter of the local user according to a preset step size control parameter, a preset weight of the feature change speed, the feature distance and the feature change speed; An estimated click rate of the recommended information of the recommended topic is determined according to the characteristic information of the recommended topic and the updated observation parameter of the local user.

3. The method according to claim 1, characterized in that The determining the target user from the global users according to the estimated click rate includes: Determining an estimated engagement level of each user in the global users according to the estimated click rate and a preset engagement level; According to the estimated participation degree of each user in the global users, at least one user with the highest estimated participation degree is selected from the global users as the target user.

4. The method according to claim 1, characterized in that: Before determining the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user, the method further includes: Acquiring initial feature information of the recommended topic; The initial feature information is mapped to obtain feature information of the recommended topic, wherein the dimension of the feature information of the recommended topic is greater than the dimension of the initial feature information.

5. The method according to claim 4, characterized in that Before mapping the initial feature information to obtain the feature information of the recommended topic, the method further includes: Normalizing the initial feature information to obtain normalized initial feature information; The mapping of the initial characteristic information to obtain the characteristic information of the recommended topic includes: The normalized initial feature information is mapped to obtain feature information of the recommended topic.

6. An information push device, characterized in that: include: A click rate determination module, used to determine the estimated click rate of the recommended information of the recommended topic according to the characteristic information of the recommended topic and the characteristic information of the global user; A user determination module, used to determine a target user from the global users according to the estimated click rate; A sending module, used for sending the recommendation information of the recommended topic to the target user; The click-through rate determination module is used to randomly select multiple users from the global users as local users, and determine the characteristic distance between the characteristic information of the recommended topic and the characteristic information of each user in the local users, as well as the characteristic change speed between the characteristic information of the recommended topic and the characteristic information of each user in the local users by trajectory tracking, and determine the observation parameters of the local users based on the characteristic distance and the characteristic change speed, the observation parameters are used to represent the traffic resources corresponding to each user in the local users, and determine the estimated click-through rate of the recommended information of the recommended topic based on the characteristic information of the recommended topic and the observation parameters of the local users.

7. A server, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the server is running, the processor and the storage medium communicate through the bus, and the processor executes the machine-readable instructions to perform the steps of the information push method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the information push method according to any one of claims 1 to 5 are executed.

Images