Method, apparatus, device, and medium for determining upper limit of resource flow

Abstract

The present disclosure provides a method, apparatus, device, and medium for determining a resource traffic upper limit, relating to the field of Internet technologies, and particularly to big data technologies. The specific implementation solution is as follows: obtaining an estimated click-through rate after exposure of different users to a target resource within a set time window; calculating a target click-through rate of the target resource according to the overall click-through rate of each resource; and determining the traffic upper limit of the target resource within the time window according to the estimated click-through rate corresponding to different users and the target click-through rate. The present disclosure can determine a more accurate resource traffic upper limit.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technologies, and more particularly to big data technologies, and specifically to a method, apparatus, device, and medium for determining a resource traffic upper limit. Background Art

[0002] In the traffic purchase business of a recommendation system, it is necessary to estimate the traffic upper limit of each resource in order to better apply a traffic allocation algorithm to allocate resources.

[0003] In the prior art, the traffic upper limit of a newly released resource of an author is usually determined based on the traffic distribution of the author's historical resources. However, the estimated value obtained by this method is not accurate, thus affecting subsequent resource allocation. Summary of the Invention

[0004] The present disclosure provides a method, apparatus, device, and medium for determining a resource traffic upper limit.

[0005] According to one aspect of the present disclosure, there is provided a method for determining a resource traffic upper limit, including:

[0006] Obtaining a predicted click-through rate of different users on a target resource after exposure within a set time window;

[0007] Calculating a target click-through rate of the target resource according to the overall click-through rate of each resource;

[0008] Determining the traffic upper limit of the target resource within the time window according to the predicted click-through rate corresponding to different users and the target click-through rate.

[0009] According to another aspect of the present disclosure, there is provided a device for determining a resource traffic upper limit, including:

[0010] A predicted click-through rate obtaining module, configured to obtain a predicted click-through rate of different users on a target resource after exposure within a set time window;

[0011] A target click-through rate calculating module, configured to calculate a target click-through rate of the target resource according to the overall click-through rate of each resource;

[0012] A traffic upper limit determining module, configured to determine the traffic upper limit of the target resource within the time window according to the predicted click-through rate corresponding to different users and the target click-through rate.

[0013] According to another aspect of the present disclosure, there is provided an electronic device, including:

[0014] At least one processor; and

[0015] A memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for determining the upper limit of resource traffic according to any embodiment of the present disclosure.

[0017] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method for determining the upper limit of resource traffic according to any embodiment of the present disclosure.

[0018] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0020] Figure 1 is a flowchart of a method for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0021] Figure 2 is a flowchart of another method for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0022] Figure 3 is a flowchart of another method for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0023] Figure 4 is a flowchart of another method for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0024] Figure 5 is a framework diagram of a method for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0025] Figure 6 is a schematic diagram of a device for determining the upper limit of resource traffic according to an embodiment of the present disclosure;

[0026] Figure 7 is a block diagram of an electronic device for implementing the method for determining the upper limit of resource traffic according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0028] Figure 1 FIG. 4 is a flowchart of a method for determining the upper limit of resource traffic according to an embodiment of the present disclosure. This embodiment is applicable to the situation of estimating the upper limit of traffic for each resource in a recommendation system, and relates to the field of Internet technology, especially big data technology. This method can be executed by a device for determining the upper limit of resource traffic, which is implemented in software and / or hardware, preferably configured in an electronic device, such as a computer device or a server, etc. As Figure 1 shown, the method specifically includes the following:

[0029] S101. Obtain the predicted click-through rate values of different users on the target resource after exposure within a set time window.

[0030] S102. Calculate the target click-through rate of the target resource according to the overall click-through rate of each resource.

[0031] S103. Determine the upper limit of the traffic of the target resource within the time window according to the predicted click-through rate values corresponding to different users and the target click-through rate.

[0032] Among them, the target resource can be any resource published by any author on the platform of the recommendation system, and can include various resource types such as graphic resources or audio-visual resources. The recommendation system recommends each resource to different users according to the traffic allocation algorithm, and the users can choose to click and view the resource according to their own interests, or not click on the resource. The upper limit of the traffic of a resource represents the maximum number of users who click or access the resource. Only by accurately estimating the upper limit of resource traffic can the traffic allocation algorithm be better applied.

[0033] Generally, the recall model in the recommendation system first recalls the matching resources, and then the sorting model sorts them, and the recalled resources are exposed to the users in order. Among them, the sorting model specifically estimates the click-through rate of each resource by the user, that is, the probability that the user clicks on the resource after the resource is exposed, and then sorts each recalled resource according to the estimated situation of the click-through rate. The estimated click-through rate is also called the sorting score and is represented by the q value. The method for the sorting model to estimate the q value can be implemented by any existing technology, and the present disclosure does not make any limitation on this.

[0034] In an embodiment of the present disclosure, a time window is preset first, for example, 1 hour, and then the predicted click-through rate values of different users for the target resource after exposure within this time window are obtained. After the target resource is exposed, the q value of the target resource can be updated every other time window, and the traffic limit of the target resource is calculated based on the q value.

[0035] Specifically, first calculate the target click-through rate of the target resource according to the overall click-through rate of each resource. Among them, the overall click-through rate is calculated based on the predicted click-through rate of each resource by the ranking model. For example, it can be the average of the click-through rates of each resource, that is, the overall click-through rate of the recommendation system. The target click-through rate represents the lowest acceptable click-through rate for the target resource. By calculating the target click-through rate of the target resource based on the overall click-through rate of each resource, the calculated target click-through rate can have a certain factual basis and be more real and accurate.

[0036] After having the lowest acceptable click-through rate for the target resource, the traffic limit of the target resource within the time window can be determined according to the predicted click-through rate values corresponding to different users and the target click-through rate. Among them, the target click-through rate can generally represent the lowest acceptable click-through rate for the target resource. Therefore, first obtain the predicted click-through rate values corresponding to different users, and then calculate the overall click-through rate based on these predicted click-through rate values. When the overall click-through rate meets the requirements of the target click-through rate, the corresponding number of users can be used as the traffic limit of the target resource within the time window.

[0037] The technical solution of the embodiment of the present disclosure is to determine the predicted traffic limit for the target resource based on the click-through rate q values of different users for the target resource predicted by the ranking model with the target click-through rate as a constraint within the set time window, and its calculation accuracy is higher. In addition, the technical solution of the embodiment of the present disclosure can also update the traffic limit of the target resource within each time window with the time window as a cycle, so that the predicted traffic limit is not a fixed value, realizing the prediction of the entire life cycle of the target resource.

[0038] Figure 2 It is a schematic flowchart of another method for determining the resource traffic limit according to an embodiment of the present disclosure. This embodiment is further optimized on the basis of the above embodiment. As Figure 2 shown, the method specifically includes the following:

[0039] S201. Obtain the predicted click-through rate values of different users for the target resource after exposure within the set time window.

[0040] S202. Multiply the overall click-through rate of each resource by the set percentage to obtain the target click-through rate of the target resource, where the set percentage represents the lowest acceptable degree of the target click-through rate.

[0041] For example, the overall click-through rate of each resource is 0.1. When the set percentage is 80%, the target click-through rate is 0.1 * 80% = 0.08. It should be noted that the present disclosure does not limit the specific value of the set percentage, and can be configured according to actual needs and the lowest acceptable degree of the target click-through rate.

[0042] S203. Sort the predicted click-through rates corresponding to different users in ascending order.

[0043] S204. Determine the nth predicted click-through rate in the sorted order, and use n as the traffic upper limit of the target resource within the time window.

[0044] Wherein, n is a natural number and satisfies the following conditions: the average value of the first predicted click-through rate to the nth predicted click-through rate is greater than or equal to the target click-through rate, and the average value of the first predicted click-through rate to the (n + 1)th predicted click-through rate is less than the target click-through rate.

[0045] Specifically, the sorting model predicts the click-through rate of each user, and the results are different. Sort in descending order according to the size of the predicted click-through rate, that is, sort from the first user with the smallest predicted value to the Nth user with the largest predicted value, where N is a natural number representing the number of users. Use the target click-through rate as a constraint condition, that is, determine the predicted click-through rate of the nth user from it, and make the average value of the first predicted click-through rate to the nth predicted click-through rate greater than or equal to the target click-through rate, and the average value of the first predicted click-through rate to the (n + 1)th predicted click-through rate less than the target click-through rate. In this way, when the above conditions are met, the determined n is the traffic upper limit of the target resource within the time window. That is to say, the upper limit of the number of users to whom the target resource is recommended is n. If the traffic continues to be greater than n, then the overall click-through rate will be less than the target click-through rate and the preset target cannot be achieved.

[0046] The technical solution of the embodiment of the present disclosure determines the target click-through rate by multiplying the overall click-through rate of each resource by the set percentage, and then uses the target click-through rate as a constraint within the set time window, and determines the predicted traffic upper limit of the target resource according to the click-through rate q value of different users predicted by the sorting model, and its calculation accuracy is higher. In addition, the traffic upper limit of the target resource within each time window can be updated with the time window as a cycle, so that the predicted traffic upper limit is not a fixed value, and the prediction of the entire life cycle of the target resource is realized.

[0047] Figure 3 It is a schematic flowchart of another method for determining the resource traffic upper limit according to the embodiment of the present disclosure. This embodiment is further optimized on the basis of the above embodiment. As Figure 3 shown, the method specifically includes the following:

[0048] S301. Classify the target resource to obtain the target type to which the target resource belongs.

[0049] S302. According to the user's historical click behavior, obtain the first potential audience interested in the resources belonging to the target type.

[0050] S303. Use the number of the first potential audience as the upper limit of the cold start estimated traffic of the target resource.

[0051] S304. Obtain the pre - estimated click - through rates of different users on the target resource after exposure within the set time window.

[0052] S305. Calculate the target click - through rate of the target resource according to the overall click - through rates of each resource.

[0053] S306. Determine the upper limit of the traffic of the target resource within the time window according to the pre - estimated click - through rates corresponding to different users and the target click - through rate.

[0054] Specifically, before determining the upper limit of the estimated traffic of the target resource according to the pre - estimated click - through rate q value of different users on the target resource, it is necessary to calculate after the target resource has been distributed for a period of time. Then, in the initial stage of the distribution of the target resource, the embodiments of the present disclosure perform cold start estimation according to S301 - S303 above. That is, the upper limit of the cold start estimated traffic is used as the estimated result when the target resource is initially released.

[0055] In cold start estimation, the embodiments of the present disclosure use the number of potential audiences of the target resource as the upper limit of the traffic of the target resource. Among them, potential audiences refer to users who may be interested in the target resource. Specifically, according to the user's historical click behavior, users interested in the resources belonging to the target type are determined and used as the first potential audience. The target type is obtained by classifying the target resource. For example, if the target resource is a video introducing outdoor hiking supplies, then its target type can be classified as outdoor equipment. It should be noted that all possible types to which resources can belong can be pre - statistically configured according to a database or a knowledge base, and the resource classification method can be implemented by any existing technical method such as text classification or label classification. The present disclosure does not make any limitation in this regard.

[0056] The technical solution of the embodiments of the present disclosure, before accurately determining the upper limit of the estimated traffic of the target resource according to the click - through rate q value of different users on the target resource and the target click - through rate, in the initial stage of the distribution of the target resource, that is, the cold start stage, obtains the upper limit of the cold start estimated traffic based on the potential audience of the target resource, further realizing the estimation of the traffic upper limit in the complete life cycle.

[0057] Figure 4It is a schematic flowchart of another method for determining the upper limit of resource traffic according to an embodiment of the present disclosure. This embodiment is further optimized based on the above embodiment. As Figure 4 shown, the method specifically includes the following:

[0058] S401. Classify the target resource to obtain the target type to which the target resource belongs.

[0059] S402. According to the historical click behavior of the user, obtain the first potential audience interested in the resources belonging to the target type.

[0060] S403. Obtain at least one point-of-interest information under the target type.

[0061] For example, when the target type of the target resource is classified as outdoor equipment, the points of interest can further be determined as tents, sleeping bags, backpacks, moisture-proof pads, etc. under the type of outdoor equipment.

[0062] S404. According to the historical click behavior of the user, obtain the second potential audience interested in the resources including at least one point-of-interest information.

[0063] S405. Take the fans of the author who publishes the target resource as the third potential audience of the target resource.

[0064] S406. Take the implicit fans of the author who publishes the target resource as the fourth potential audience of the target resource, where the implicit fans refer to those who have not followed the author and the overall click-through rate of the implicit fans on the resources published by the author is higher than the average click-through rate of the author's fans.

[0065] S407. Add the numbers of the first potential audience, the second potential audience, the third potential audience, and the fourth potential audience to obtain the upper limit of the cold start estimated traffic of the target resource.

[0066] S408. Obtain the pre-estimated click-through rate values after the exposure of the target resource by different users within the set time window.

[0067] S409. Calculate the target click-through rate of the target resource according to the overall click-through rate of each resource.

[0068] S410. Determine the upper limit of the traffic of the target resource within the time window according to the pre-estimated click-through rate values corresponding to different users and the target click-through rate.

[0069] Specifically, the potential audiences in this embodiment not only refer to the first potential audiences who are interested in resources belonging to the target type, but also include the second potential audiences who are interested in resources including at least one interest point information, the third potential audiences who are fans of the author, and the fourth potential audiences who are implicit fans of the author. Among them, implicit fans refer to those who have not followed the author, but their overall click-through rate on the resources published by the author is higher than the average click-through rate of the author's fans. Therefore, in the technical solution of the embodiment of the present disclosure, the potential audiences of the target resources are determined from four aspects, and their numbers are added together to obtain the upper limit of the cold start estimated traffic. Thus, a more accurate upper limit of the cold start estimated traffic can be determined based on a more comprehensive and multi-dimensional potential audience, and the upper limit of the traffic of the target resource can be further estimated throughout the entire life cycle.

[0070] Figure 5 is a framework diagram of another method for determining the upper limit of resource traffic according to an embodiment of the present disclosure. As Figure 5 shown, the resources published by the author will be processed in the material library, and information such as the type and interest points of the resources will be obtained by classification. At the same time, the resources processed by the material library will enter the recommendation system for distribution, and the recall model will recall the matching resources for the users, and the sorting model will sort the recalled resources, and finally be exposed to the users. In the traffic estimation system, first, the cold start estimation of the traffic upper limit of the resources is performed according to information such as the type and interest points of the resources, and then, based on the log of the sorting scores output by the sorting model, the upper limit of the traffic of the target resource is accurately estimated based on the click-through rate q value and the target click-through rate of different users on the target resource, and the traffic upper limit will also be updated in each time window. In this way, a more accurate and full life cycle upper limit of resource traffic estimation is achieved.

[0071] Figure 6 is a schematic diagram of a device for determining the upper limit of resource traffic according to an embodiment of the present disclosure. This embodiment is applicable to the situation of estimating the upper limit of the traffic of each resource in the recommendation system, and relates to the field of Internet technology, especially big data technology. This device can implement the method for determining the upper limit of resource traffic described in any embodiment of the present disclosure. As Figure 6 shown, the device 600 specifically includes:

[0072] A click-through rate pre-estimation value acquisition module 601, configured to acquire the pre-estimation value of the click-through rate of different users on the target resource after exposure within a set time window;

[0073] A target click-through rate calculation module 602, configured to calculate the target click-through rate of the target resource according to the overall click-through rate of each resource;

[0074] A traffic upper limit determination module 603, configured to determine the traffic upper limit of the target resource within the time window according to the pre-estimation value of the click-through rate corresponding to different users and the target click-through rate.

[0075] Optionally, the target click-through rate calculation module 602 is specifically configured to:

[0076] Multiply the overall click-through rate of each resource by a set percentage to obtain the target click-through rate of the target resource, where the set percentage represents the lowest acceptable degree of the target click-through rate.

[0077] Optionally, the traffic upper limit determination module 603 includes:

[0078] A sorting unit, configured to sort the estimated click-through rates corresponding to different users in ascending order;

[0079] A first traffic upper limit determination unit, configured to determine the nth estimated click-through rate in the sorted order and use n as the traffic upper limit of the target resource within the time window;

[0080] where n is a natural number and satisfies the following conditions: the average value of the first estimated click-through rate to the nth estimated click-through rate is greater than or equal to the target click-through rate, and the average value of the first estimated click-through rate to the (n + 1)th estimated click-through rate is less than the target click-through rate.

[0081] Optionally, the device further includes:

[0082] An update module, configured to update the traffic upper limit of the target resource within each time window with the time window as a period.

[0083] Optionally, the device further includes a cold start estimation module, and the cold start estimation module includes:

[0084] A classification unit, configured to classify the target resource to obtain the target type to which the target resource belongs before the click-through rate estimation value acquisition module 601 acquires the estimated click-through rates of different users on the target resource after exposure within a set time window;

[0085] A first potential audience acquisition unit, configured to acquire a first potential audience interested in resources belonging to the target type according to the historical click behavior of users;

[0086] A second traffic upper limit determination unit, configured to use the number of the first potential audience as the traffic upper limit for cold start estimation of the target resource.

[0087] Optionally, the cold start estimation module further includes:

[0088] The second potential audience acquisition unit is configured to acquire at least one point of interest information under the target type; and acquire a second potential audience interested in resources including the at least one point of interest information according to the user's historical click behavior.

[0089] Correspondingly, the second traffic ceiling determination unit is specifically configured to:

[0090] Add the quantities of the first potential audience and the second potential audience to obtain the cold start estimated traffic ceiling of the target resource.

[0091] Optionally, the cold start estimation module further includes:

[0092] A third potential audience acquisition unit, configured to use the fans of the author who publishes the target resource as the third potential audience of the target resource;

[0093] Correspondingly, the second traffic ceiling determination unit is specifically configured to:

[0094] Add the quantities of the first potential audience, the second potential audience, and the third potential audience to obtain the cold start estimated traffic ceiling of the target resource.

[0095] Optionally, the cold start estimation module further includes:

[0096] A fourth potential audience acquisition unit, configured to use the implicit fans of the author who publishes the target resource as the fourth potential audience of the target resource, where the implicit fans refer to those who have not followed the author and whose overall click-through rate on the resources published by the author is higher than the average click-through rate of the author's fans;

[0097] Correspondingly, the second traffic ceiling determination unit is specifically configured to:

[0098] Add the quantities of the first potential audience, the second potential audience, the third potential audience, and the fourth potential audience to obtain the cold start estimated traffic ceiling of the target resource.

[0099] Optionally, the cold start estimated traffic ceiling is used as the estimated result when the target resource is initially published.

[0100] The above product can execute the method provided by any embodiment of the present disclosure, and has corresponding functional modules and beneficial effects for executing the method.

[0101] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information and other processes all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0102] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0103] Figure 7 FIG. shows a schematic block diagram of an exemplary electronic device 700 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0104] As Figure 7 shown, the device 700 includes a computing unit 701 that can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0105] A plurality of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a magnetic disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0106] The computing unit 701 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 executes the various methods and processes described above, such as the method for determining the resource traffic upper limit. For example, in some embodiments, the method for determining the resource traffic upper limit can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the method for determining the resource traffic upper limit described above can be executed. Alternatively, in other embodiments, the computing unit 701 can be configured to execute the method for determining the resource traffic upper limit by any other suitable means (e.g., by means of firmware).

[0107] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0108] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program code is executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0109] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0110] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic, speech, or tactile input).

[0111] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0112] A computer system may include a client and a server. The client and the server are generally far from each other and usually interact via a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services. The server can also be a server of a distributed system or a server combined with blockchain.

[0113] Artificial intelligence is a discipline that studies how to make a computer simulate certain human thinking processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.), and it has both hardware-level technologies and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include several major directions such as computer vision technology, speech recognition technology, natural language processing technology, machine learning / deep learning technology, big data processing technology, and knowledge graph technology.

[0114] Cloud computing refers to a technical system that accesses an elastic and scalable shared physical or virtual resource pool through a network. The resources can include servers, operating systems, networks, software, applications, and storage devices, etc., and the resources can be deployed and managed in a on-demand and self-service manner. Through cloud computing technology, it can provide efficient and powerful data processing capabilities for the application and model training of technologies such as artificial intelligence and blockchain.

[0115] It should be understood that various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions provided in this disclosure can be achieved, and no limitation is imposed herein.

[0116] The above specific implementation manners do not constitute a limitation to the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for determining the upper limit of resource traffic, which is used in the traffic prediction system of a recommendation system. The recommendation system also includes a recall model and a ranking model. The recall model is used to recall matching resources, and the ranking model is used to estimate the click-through rate of each resource for a user and rank the recalled resources according to the estimated click-through rate. The method includes: Classify the target resource to obtain the target type to which the target resource belongs; According to the user's historical click behavior, obtain the first potential audience interested in resources belonging to the target type; Obtain at least one point-of-interest information under the target type; According to the user's historical click behavior, obtain the second potential audience interested in resources including the at least one point-of-interest information; Take the fans of the author who publishes the target resource as the third potential audience of the target resource; Take the implicit fans of the author who publishes the target resource as the fourth potential audience of the target resource, where the implicit fans refer to those who have not followed the author and the overall click-through rate of the implicit fans for the resources published by the author is higher than the average click-through rate of the author's fans; Add the numbers of the first potential audience, the second potential audience, the third potential audience, and the fourth potential audience to obtain the upper limit of the cold-start estimated traffic of the target resource; According to the click-through rate logs output by the ranking model, obtain the pre-estimated click-through rates of different users for the target resource after exposure within a set time window; Multiply the overall click-through rate of each resource by a set percentage to obtain the target click-through rate of the target resource, where the overall click-through rate represents the overall click-through rate of the recommendation system, and the target click-through rate represents the lowest acceptable click-through rate for the target resource; Sort in descending order according to the magnitudes of the pre-estimated click-through rates corresponding to different users; Determine the nth pre-estimated click-through rate in the sorted order and take n as the upper limit of the traffic of the target resource within the time window, where the upper limit of the traffic represents the maximum number of users who click on or access the target resource; n is a natural number and satisfies the following conditions: the average value of the first pre-estimated click-through rate to the nth pre-estimated click-through rate is greater than or equal to the target click-through rate, and the average value of the first pre-estimated click-through rate to the (n + 1)th pre-estimated click-through rate is less than the target click-through rate; Taking the time window as a period, update the upper limit of the traffic of the target resource within each time window.

2. The method according to claim 1, wherein, The set percentage represents the lowest acceptable degree of the target click-through rate.

3. The method according to claim 1, wherein, The upper limit of the cold-start estimated traffic is used as the estimated result when the target resource is initially published.

4. A device for determining the upper limit of resource traffic, which is configured in the traffic prediction system of a recommendation system. The recommendation system also includes a recall model and a ranking model. The recall model is used to recall matching resources, and the ranking model is used to estimate the click-through rate of each resource for a user and rank the recalled resources according to the estimated click-through rate. The device includes: A click-through rate prediction value acquisition module, configured to obtain the predicted click-through rate values of different users for the target resource after exposure within a set time window according to the click-through rate logs output by the sorting model; A target click-through rate calculation module, configured to calculate the target click-through rate of the target resource according to the overall click-through rate of each resource; A traffic upper limit determination module, configured to determine the traffic upper limit of the target resource within the time window according to the predicted click-through rate values corresponding to different users and the target click-through rate, where the traffic upper limit represents the maximum number of users who click on or access the target resource; Among them, the target click-through rate calculation module is specifically configured to: Multiply the overall click-through rate of each resource by a set percentage to obtain the target click-through rate of the target resource, where the overall click-through rate represents the overall click-through rate of the recommendation system, and the target click-through rate represents the lowest acceptable click-through rate for the target resource; The traffic upper limit determination module includes: A sorting unit, configured to perform a descending order sorting according to the magnitudes of the predicted click-through rate values corresponding to different users; A first traffic upper limit determination unit, configured to determine the nth predicted click-through rate value in the sorted order and use n as the traffic upper limit of the target resource within the time window; Among them, n is a natural number and satisfies the following conditions: the average value of the first predicted click-through rate value to the nth predicted click-through rate value is greater than or equal to the target click-through rate, and the average value of the first predicted click-through rate value to the (n + 1)th predicted click-through rate value is less than the target click-through rate; The device further includes a cold start prediction module, and the cold start prediction module includes: A classification unit, configured to classify the target resource to obtain the target type to which the target resource belongs before the click-through rate prediction value acquisition module obtains the predicted click-through rate values of different users for the target resource after exposure within a set time window; A first potential audience acquisition unit, configured to obtain a first potential audience interested in resources belonging to the target type according to the historical click behaviors of users; A second potential audience acquisition unit, configured to obtain at least one interest point information under the target type; and obtain a second potential audience interested in resources including the at least one interest point information according to the historical click behaviors of users; A third potential audience acquisition unit, configured to use the fans of the author who publishes the target resource as the third potential audience of the target resource; A fourth potential audience acquisition unit, configured to use the implicit fans of the author who publishes the target resource as the fourth potential audience of the target resource, where the implicit fans refer to those who have not followed the author, and the overall click-through rate of the implicit fans for the resources published by the author is higher than the average click-through rate of the author's fans; A second traffic upper limit determination unit, configured to add the numbers of the first potential audience, the second potential audience, the third potential audience, and the fourth potential audience to obtain the cold start predicted traffic upper limit of the target resource; The device further includes an update module, configured to update the traffic upper limit of the target resource within each time window with the time window as a period.

5. An electronic device, Comprising: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for determining the upper limit of resource traffic according to any one of claims 1-3.

6. A non-transitory computer-readable storage medium storing computer instructions, Wherein, The computer instructions are used to cause a computer to execute the method for determining the upper limit of resource traffic according to any one of claims 1-3.

7. A computer program product comprising a computer program which, when executed by a processor, implements the method for determining the upper limit of resource traffic according to any one of claims 1-3.

Images