Content pushing method and related equipment

Through the content push method for news information, the target content and push object group identification are determined, candidate devices are recalled and equipment allocation is performed, the problem of slow news push speed in the existing technology is solved, and fast issuance and efficient push is achieved.

CN119025747APending Publication Date: 2024-11-26TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310606215.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-25
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing content push method is slow when quickly distributing news information, affecting user click-through rates, and when multiple content distribution clients are installed, the push speed is difficult to optimize.

Method used

By determining the target content and push object group identification, recalling the candidate push device collection, and quickly allocating candidate push devices to the target machine node based on the mapping relationship between device and node information, conducting business requirements analysis to determine the target push device, and achieving rapid issuance .

Benefits of technology

While ensuring the accuracy of push, it greatly improves the speed and efficiency of content push, and can complete the first recommendation of hot spots and emergencies in one minute.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025747A_ABST
    Figure CN119025747A_ABST
Patent Text Reader

Abstract

The invention discloses a content pushing method and related equipment. The method comprises the following steps: recalling candidate push equipment corresponding to target content based on a push object group identifier corresponding to the target content to be pushed and an equipment set corresponding to a preset object group identifier; for each candidate pushing device, determining a target machine node used for processing the candidate pushing device from the machine nodes based on an association relationship of mapping points of the device information of the candidate pushing device and the node information of the machine nodes in a preset data space, distributing the candidate pushing equipment to the target machine node; through the target machine node, performing service demand analysis on candidate pushing equipment allocated to the target machine node, so as to determine target pushing equipment from the allocated candidate pushing equipment; and pushing the target content to the target pushing equipment. According to the method and the device, the content pushing speed and efficiency can be greatly improved while the pushing accuracy is ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a content push method and related equipment. Background Art

[0002] With the rapid development of Internet technology, multimedia content on the Internet has increased explosively. Users obtain multimedia content mainly through two ways: pull and push. The first way requires users to actively log in to the content distribution client and check the content stream (feed stream); the second way is to actively send push messages to users who may be interested when there is a hot spot or content that users are interested in. Therefore, push is an important way to distribute content.

[0003] Among the current related technologies, the industry mainly focuses on the research related to personalized push, that is, the optimization and implementation of push strategies, such as how an article can find its target audience. Personalized push content does not have particularly high requirements for timeliness, as long as it can guarantee a basic success rate of reaching. However, for some content with high "timeliness", such as news information, it is different from ordinary personalized content push. The most important requirement for news information is to achieve fast delivery. The push speed also affects the user's click to pull up, because the average user's mobile phone will have multiple content distribution clients installed. For a hot news push, the user will click on the content distribution client that pushes the news first, and will ignore the push of other content distribution clients after reading it; therefore, the delivery speed is also an important optimization goal for news push. The current content push method has a relatively slow content delivery speed. Summary of the invention

[0004] The embodiments of the present application provide a content push method and related equipment. The related equipment may include a content push device, an electronic device, a computer-readable storage medium and a computer program product, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy.

[0005] The present application provides a content push method, including:

[0006] Determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content;

[0007] Based on the push object group identifier and the device set corresponding to the preset object group identifier, recall at least one push device set corresponding to the target content, the push device set including at least one candidate push device;

[0008] For each candidate push device, based on the association between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine a target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node;

[0009] By means of the target machine node, a business demand analysis is performed on the candidate push devices assigned to the target machine node, so as to determine the target push device from the assigned candidate push devices;

[0010] The target content is pushed to the target push device.

[0011] Accordingly, an embodiment of the present application provides a content push device, including:

[0012] A determination unit, used to determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content;

[0013] A recall unit, configured to recall at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier, wherein the push device set includes at least one candidate push device;

[0014] a node determination unit, configured to determine, for each candidate push device, a target machine node for processing the candidate push device from among the machine nodes based on an association relationship between device information of the candidate push device and mapping points of node information of each machine node in a preset data space, and assign the candidate push device to the target machine node;

[0015] An analyzing unit, configured to perform a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node, so as to determine a target push device from the assigned candidate push devices;

[0016] A pushing unit is used to push the target content to the target pushing device.

[0017] Optionally, in some embodiments of the present application, the content push apparatus may further include a device set determination unit as follows:

[0018] The device set determination unit is used to obtain interaction data, where the interaction data includes interaction attribute information of at least one interactive object; based on the interaction attribute information of each interactive object, the devices using the interactive object are aggregated to obtain at least one original device set; for each original device set, based on at least one preset object group identifier, the original device set is segmented to obtain device sets corresponding to each preset object group identifier.

[0019] Optionally, in some embodiments of the present application, the node determination unit may include a first mapping subunit, a second mapping subunit and a node determination subunit, as follows:

[0020] The first mapping subunit is used to map the candidate push device to a preset data space based on the device information of the candidate push device, so as to obtain a device mapping point of the candidate push device in the preset data space;

[0021] A second mapping subunit is used to map each machine node to the preset data space based on the node information of each machine node, so as to obtain a node mapping point of each machine node in the preset data space;

[0022] The node determination subunit is used to search for a target machine node for processing the candidate push device from the machine nodes according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space.

[0023] Optionally, in some embodiments of the present application, the preset data space is a hash space ring composed of hash values;

[0024] The first mapping subunit may be specifically configured to perform a hash operation on the device information of the candidate push device to obtain a first hash value; map the candidate push device to the hash space ring according to the first hash value to obtain a device mapping point of the candidate push device on the hash space ring;

[0025] The second mapping subunit can be specifically used to perform hash operation on the node information of each machine node to obtain a second hash value; according to the second hash value, map each machine node to the hash space ring to obtain a node mapping point of each machine node on the hash space ring.

[0026] Optionally, in some embodiments of the present application, the node determination subunit can be specifically used to start from the position corresponding to the device mapping point, and perform node search along the hash space ring in a preset direction; when the node mapping point is found, the machine node corresponding to the node mapping point is determined as the target machine node for processing the candidate push device.

[0027] Optionally, in some embodiments of the present application, the content push device may further include a standby node replacement unit, as follows:

[0028] The standby node replacement unit is configured to obtain at least one standby machine node of the target machine node when an abnormality occurs in the target machine node; divide the candidate push devices allocated to the target machine node based on the number of the standby machine nodes to obtain at least one device subset, wherein the device subset includes at least one candidate push device; allocate the at least one device subset to each standby machine node, and determine each standby machine node as a new target machine node for processing the allocated device subset.

[0029] Optionally, in some embodiments of the present application, the device set corresponding to the preset object group identifier is stored in a remote server;

[0030] The content push device may further include a cache unit, as follows:

[0031] The cache unit is used to cache the device set corresponding to the preset object group identifier stored in the remote server to the preset service storage cluster, so that at least two service nodes in the preset service storage cluster cache the device set corresponding to the preset object group identifier;

[0032] The recall unit may include a reading subunit, a re-reading subunit and a determining subunit, as follows:

[0033] The reading subunit is used to read the device set corresponding to the push object group identifier through a target service node, and the target service node is any service node;

[0034] A re-reading subunit is used for, when reading is abnormal, performing an identification query in the reference service node based on the push object group identifier through a reference service node to read the device set corresponding to the push object group identifier, wherein the reference service node is a service node other than the target service node;

[0035] The determination subunit is configured to determine the read device set as at least one push device set corresponding to the target content.

[0036] Optionally, in some embodiments of the present application, the content push device may further include a version detection unit, as follows:

[0037] The version detection unit is used to detect the version information of the device set stored in each service node; when it is detected that the version information of the device set in the service node is inconsistent, the service node is triggered to send a data synchronization instruction to the remote server; based on the data synchronization instruction, the device set currently stored in the remote server is synchronized to each service node.

[0038] Optionally, in some embodiments of the present application, the content push device may further include a priority unit as follows:

[0039] The priority unit is used to determine the sensitivity of the target content to the push delay according to the content attribute information of the target content; determine the priority information of the target content based on the sensitivity, add the push task corresponding to the target content to the message queue corresponding to the priority information; and take out the push task of the target content from the message queue corresponding to the priority information;

[0040] The node determination unit may be specifically configured to allocate the candidate push device to the target machine node based on the push task of the target content.

[0041] Optionally, in some embodiments of the present application, the analysis unit may include a logic block determination subunit, a hash subunit, a third mapping subunit, and a filtering subunit, as follows:

[0042] The logic block determination subunit is used to determine, for each candidate push device assigned to the target machine node, a push record logic block of the candidate push device, wherein the push record logic block includes at least one bit, and state identification information of the bit represents interaction information corresponding to the candidate push device and the content;

[0043] A hash subunit, used to perform a hash operation on the target content to obtain a content hash value;

[0044] a third mapping subunit, configured to map the target content to the push record logic block based on the content hash value, and determine a mapping position of the target content in the push record logic block;

[0045] The filtering subunit is used to filter the candidate push devices according to the status identification information of the bit corresponding to the mapping position to obtain the target push device.

[0046] Optionally, in some embodiments of the present application, the filtering sub-unit can be specifically used to filter out the candidate push device when the status identification information of each bit corresponding to the mapping position indicates that each bit has been set; when the status identification information of each bit corresponding to the mapping position indicates that there is a bit that is not set among the bits, retain the candidate push device and update the status identification information of the bit to be set, so as to obtain the target push device corresponding to the target machine node.

[0047] Optionally, in some embodiments of the present application, the analysis unit may include a data acquisition subunit and a device determination subunit, as follows:

[0048] The data acquisition subunit is used to acquire, for each candidate push device assigned to the target machine node, historical push data of the candidate push device in at least one dimension through the target machine node;

[0049] The device determination subunit is used to determine the candidate push device as the target push device when the historical push data in the at least one dimension meets a preset condition.

[0050] An electronic device provided in an embodiment of the present application includes a processor and a memory, wherein the memory stores a plurality of instructions, and the processor loads the instructions to execute the steps in the content push method provided in the embodiment of the present application.

[0051] The embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps in the content push method provided in the embodiment of the present application.

[0052] In addition, an embodiment of the present application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the steps in the content push method provided in the embodiment of the present application.

[0053] The embodiment of the present application provides a content push method and related equipment; the target content to be pushed and the push object group identifier corresponding to the target content can be determined; and at least one machine node for pushing the target content can be determined; based on the device set corresponding to the push object group identifier and the preset object group identifier, at least one push device set corresponding to the target content is recalled, and the push device set includes at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, a target machine node for processing the candidate push device is determined from the machine nodes, and the candidate push device is assigned to the target machine node; through the target machine node, a business demand analysis is performed on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and the target content is pushed to the target push device. When triggering the push of target content, the present application can determine the corresponding push device set based on the push object group identifier specified by the target content, without having to find the push audience object based on the object attribute information of massive candidate objects during the content push process, which greatly improves the speed of determining the candidate push devices; in addition, the candidate push devices can be quickly allocated and sent to the corresponding machine nodes based on the association relationship between the mapping points of the candidate push devices and the machine nodes in the preset data space, thereby achieving load balancing of each machine node, and thus performing business demand analysis on the allocated candidate push devices through the machine nodes, so as to push content to the devices that really need to be pushed, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0055] Figure 1a It is a scenario diagram of the content push method provided in an embodiment of the present application;

[0056] Figure 1b is a flow chart of a content push method provided by an embodiment of the present application;

[0057] Figure 1c It is a system framework diagram of the content push method provided by the embodiment of the present application;

[0058] Figure 1d is an illustration of a content push method provided in an embodiment of the present application;

[0059] Figure 2 is another flow chart of the content push method provided by an embodiment of the present application;

[0060] Figure 3 It is a structural diagram of a content push device provided in an embodiment of the present application;

[0061] Figure 4 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0063] The embodiment of the present application provides a content push method and related equipment, and the related equipment may include a content push device, an electronic device, a computer-readable storage medium and a computer program product. The content push device may be integrated in an electronic device, and the electronic device may be a terminal or a server.

[0064] It is understandable that the content push method of this embodiment can be executed on the terminal, can be executed on the server, or can be executed by the terminal and the server together. The above examples should not be understood as limiting the present application.

[0065] like Figure 1a As shown, a terminal and a server jointly execute a content push method as an example. The content push system provided in the embodiment of the present application includes a terminal 10 and a server 11, etc. The terminal 10 and the server 11 are connected via a network, such as a wired or wireless network connection, etc., wherein the content push device can be integrated in the server.

[0066] The server 11 can be used to: determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; based on the device set corresponding to the push object group identifier and the preset object group identifier, recall at least one push device set corresponding to the target content, the push device set including at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; through the target machine node, perform business demand analysis on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and push the target content to the target push device. The server 11 can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0067] The terminal 10 can be used to receive the target content pushed by the server 11 through the target push device (specifically, the terminal 10), and display the target content on the content distribution client installed in the target push device. The terminal 10 can include a mobile phone, an intelligent voice interaction device, a smart home appliance, a vehicle terminal, an aircraft, a tablet computer, a laptop computer, or a personal computer (PC). A client can also be set on the terminal 10, and the client can be an application client or a browser client, etc.

[0068] The content push method provided in the embodiment of the present application relates to the cloud network direction and the big data direction in the field of cloud technology.

[0069] Among them, cloud technology refers to a hosting technology that unifies hardware, software, network and other resources in a wide area network or local area network to realize data computing, storage, processing and sharing. Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, application technology, etc. based on the cloud computing business model application. It can form a resource pool, which is used on demand and flexible and convenient. Cloud computing technology will become an important support. The background services of the technical network system require a large amount of computing and storage resources, such as video websites, picture websites and more portal websites. With the high development and application of the Internet industry, in the future, each item may have its own identification mark, and all need to be transmitted to the background system for logical processing. Data of different levels will be processed separately. All kinds of industry data require strong system backing support, which can only be achieved through cloud computing.

[0070] Cloud web is a general term for Internet technology, integration, and applications in the cloud era. Cloud web uses the cloud to connect network resources. It can be said that independent servers are integrated into a virtual network service system through network technology, providing users with a resource pool that can be used on demand and paid for.

[0071] Big data refers to a collection of data that cannot be captured, managed, and processed by conventional software tools within a certain time frame. It is a massive, high-growth, and diversified information asset that requires new processing models to have stronger decision-making power, insight discovery, and process optimization capabilities. With the advent of the cloud era, big data has also attracted more and more attention. Big data requires special technologies to effectively process large amounts of data within a tolerable time frame. Technologies applicable to big data include large-scale parallel processing databases, data mining, distributed file systems, distributed databases, cloud computing platforms, the Internet, and scalable storage systems.

[0072] It should be noted that the description order of the following embodiments is not intended to limit the preferred order of the embodiments.

[0073] This embodiment will be described from the perspective of a content push device. The content push device may be integrated into an electronic device, which may be a server or a terminal.

[0074] It is understandable that in the specific implementation of this application, user information, such as interests and other related data, is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0075] like Figure 1b As shown, the specific process of the content push method can be as follows:

[0076] 101. Determine target content to be pushed and a push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content.

[0077] The target content is the content to be recommended, and its content type is not limited, for example, the target content can be text, audio, image, video, etc. For example, the target content can be a news article.

[0078] The push target group identifier may be the tag information of the recommended target group, specifically the audience tag of the content to be pushed. For example, the push target group identifier may be the target interest, the device used (such as the type of mobile phone), the region where the target is located, etc., which is not limited in this embodiment.

[0079] Among them, the machine node can be a server node used to schedule push tasks. Specifically, it can be used to perform business demand analysis on each assigned push task to determine the tasks that really need to be pushed. The business demand analysis may include push duplicate filtering, push frequency control, quota management, and dispersion.

[0080] Push is an important way to distribute content. Current related technologies mainly focus on personalized push and how to improve the effect of personalized push, such as how a piece of news can find its target audience. Personalized push does not have particularly high requirements on the timeliness of push, as long as a good push success rate can be guaranteed.

[0081] However, for some content with high "timeliness", such as news information, it is different from ordinary personalized content push. The most important requirement for news information is to achieve fast delivery. The delivery speed also affects the user's click-to-pull, because generally users will install multiple information / content apps (Application) on their mobile phones. For a hot news push, users will click on the app that pushes the news first, and will ignore the push of other apps after reading it. Therefore, the delivery speed is also an important optimization goal of news push. The current content push method has a relatively slow content delivery speed.

[0082] The content push method provided in this application is mainly aimed at some content pushes with high "timeliness", such as news push. News push is different from ordinary personalized content push. The biggest feature of news is "newness", and news push must also reflect timeliness. Therefore, the most important requirement for news push is to achieve rapid delivery. This application uses the architectural design of a news push system with rapid delivery as the goal, and can achieve "first push on the entire network" for the push of many hot emergencies. Specifically, this application realizes a fast delivery link for news push by optimizing the architectural design of multiple module processes such as push recall and push scheduling. For a large hot emergency push to all active users, the content push system provided by this application can complete the push within 1 minute.

[0083] In actual applications, content push can be triggered in a variety of business scenarios. For example, it can be a message push triggered by a push operation platform, a special trigger process for identifying new users, an automated trigger process that automatically pushes content to users based on their interests, or content notification push generated by user interactive operations (such as following or commenting).

[0084] 102. Based on the push object group identifier and the device set corresponding to the preset object group identifier, recall at least one push device set corresponding to the target content, where the push device set includes at least one candidate push device.

[0085] Among them, the matching degree between the push object group identifier and each preset object group identifier can be determined, and the device set of the preset object group identifier whose matching degree meets the preset matching conditions is determined as the device set corresponding to the push object group identifier. The device set corresponding to the push object group identifier is also the push device set corresponding to the target content.

[0086] The preset matching condition may be set according to actual conditions, such as the preset matching condition may be that the matching degree is greater than the preset matching degree.

[0087] The candidate push device may be a device used by the candidate push user, such as a mobile phone, a computer, etc.

[0088] Optionally, in this embodiment, the content push method may further include:

[0089] Acquiring interactive data, wherein the interactive data includes interactive attribute information of at least one interactive object;

[0090] Based on the interactive attribute information of each interactive object, aggregate the devices using the interactive object to obtain at least one original device set;

[0091] For each original device set, based on at least one preset object group identifier, the original device set is segmented to obtain device sets corresponding to each preset object group identifier.

[0092] The interactive object may be an object that interacts with the content, and the interaction here may include browsing, clicking, commenting, etc. The interactive attribute information may include preference information of the interactive object, etc.

[0093] Among them, the step of "based on the interactive attribute information of each interactive object, aggregating the usage devices of the interactive object to obtain at least one original device set" can specifically be: according to the interactive attribute information, aggregating each interactive object to obtain multiple object sets, each object set includes at least one interactive object, and different object sets may correspond to different types of interactive objects, where the type may refer to the type of interest preference, etc.; then, based on the usage devices of each interactive object, the interactive objects in the object set are replaced with their corresponding usage devices to obtain the corresponding original device set.

[0094] After obtaining the original device set, since the original device set may still contain a large amount of data, it can be cut into several small device sets according to each preset object group identifier. Segmentation of the original device set can also facilitate task parallel processing.

[0095] This embodiment can decompose and convert the push object group identifier into a specific user device list (ie, a push device set), which can greatly increase the speed of determining the push object, thereby improving the content push efficiency.

[0096] Optionally, in this embodiment, the device set corresponding to the preset object group identifier is stored in a remote server;

[0097] Before the step of “recalling at least one push device set corresponding to the target content based on the push target group identifier and the device set corresponding to the preset target group identifier”, the method may further include:

[0098] Caching the device set corresponding to the preset object group identifier stored in the remote server to the preset service storage cluster, so that at least two service nodes in the preset service storage cluster cache the device set corresponding to the preset object group identifier;

[0099] The step of “recalling at least one push device set corresponding to the target content based on the push target group identifier and the device set corresponding to the preset target group identifier” may include:

[0100] Reading a device set corresponding to the push object group identifier through a target service node, wherein the target service node is any service node;

[0101] When reading is abnormal, through a reference service node, based on the push object group identifier, an identifier query is performed in the reference service node to read the device set corresponding to the push object group identifier, and the reference service node is a service node other than the target service node;

[0102] The read device set is determined as at least one push device set corresponding to the target content.

[0103] Among them, the remote server is specifically a remote distributed file system, which can be used to store the device set corresponding to each preset object group identifier. Specifically, the remote server can regularly update the interactive attribute information of the interactive object, and according to the updated interactive attribute information, aggregate the use devices of the interactive object, thereby updating the device set corresponding to each preset object group identifier. Therefore, the device set corresponding to the preset object group identifier in the remote server can be regarded as the latest. In some embodiments, the version information of the device set in the remote server and the version information of the device set in the preset business storage cluster can be checked regularly. If the version of the remote server is newer than the version of the preset business storage cluster, data synchronization will be initiated to synchronize the device set in the remote distributed file system to the service node of the preset business storage cluster.

[0104] In order to improve the reading speed of the device set, the device set in the remote server can be cached in the service node of the preset business storage cluster corresponding to the content push system. Specifically, multiple node copies can be set to ensure the high availability of the device set stored in the preset business storage cluster. If a service node is abnormal, the corresponding device set can be obtained through other service nodes.

[0105] The reference service node may be a service node other than the target service node that caches a device set corresponding to a preset object group identifier.

[0106] Optionally, in this embodiment, the content push method may further include:

[0107] Detect the version information of the device set stored in each service node;

[0108] When it is detected that the version information of the device set in the service node is inconsistent, triggering the service node to send a data synchronization instruction to the remote server;

[0109] Based on the data synchronization instruction, the device set currently stored in the remote server is synchronized to each service node.

[0110] Among them, this embodiment needs to maintain the consistency of the version information of the device set stored in each service node. Specifically, when it is found that the version information of the device set stored in each service node in the cluster is inconsistent, a data synchronization signal can be sent to all service nodes. Based on the received data synchronization signal, all service nodes initiate a data synchronization instruction to the remote server. Based on the data synchronization instruction, the device set stored in each service node in the cluster can be refreshed to the latest data version.

[0111] Among them, optionally, in some embodiments, the content push method may also include:

[0112] When a service node synchronization failure is detected, marking the service state of the service node whose synchronization failed is an abnormal state;

[0113] The service nodes whose service status is abnormal are removed from the preset business storage cluster.

[0114] When a service node fails to synchronize data, it indicates that there may be a problem with the service node. Therefore, it can be marked as abnormal and removed to prevent upstream services from accessing the abnormal node.

[0115] 103. For each candidate push device, based on the association between the device information of the candidate push device and the mapping points of the node information of each machine node in the preset data space, determine a target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node.

[0116] Specifically, in some embodiments, each candidate push device can be mapped to a preset data space, and each machine node can also be mapped to the preset data space. Then, according to the mapping position of each machine node on the preset data space and the mapping position of each candidate push device on the preset data space, all candidate push devices are divided to obtain a set of candidate push devices assigned to each machine node, thereby assigning each candidate push device set to a corresponding machine node for processing.

[0117] Optionally, in this embodiment, the step of “determining a target machine node for processing the candidate push device from the machine nodes based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node in the preset data space” may include:

[0118] Based on the device information of the candidate push device, the candidate push device is mapped to a preset data space to obtain a device mapping point of the candidate push device in the preset data space;

[0119] Based on the node information of each machine node, each machine node is mapped to the preset data space to obtain a node mapping point of each machine node in the preset data space;

[0120] According to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space, a target machine node for processing the candidate push device is searched from the machine nodes.

[0121] The device information of the candidate push device may be a device name, etc. The node information of the machine node may be an IP (Internet Protocol Address) or a host name, etc. of the machine node.

[0122] Optionally, in this embodiment, the preset data space is a hash space ring composed of hash values;

[0123] The step of “mapping the candidate push device to a preset data space based on the device information of the candidate push device to obtain a device mapping point of the candidate push device in the preset data space” may include:

[0124] Performing a hash operation on the device information of the candidate push device to obtain a first hash value;

[0125] According to the first hash value, mapping the candidate push device to the hash space ring to obtain a device mapping point of the candidate push device on the hash space ring;

[0126] The step of “mapping each machine node to the preset data space based on the node information of each machine node to obtain a node mapping point of each machine node in the preset data space” may include:

[0127] Performing a hash operation on the node information of each machine node to obtain a second hash value;

[0128] According to the second hash value, each machine node is mapped to the hash space ring to obtain a node mapping point of each machine node on the hash space ring.

[0129] Specifically, the hash space ring can be a circular ring corresponding to the hash space of 2^32-1, that is, it is a circle composed of 2^32 points, and each point on the circle represents 1, 2, 3, ... until 2^32-1.

[0130] The hash algorithm used to perform hash operations on the device information of the candidate push device and the node information of the machine node may be the same. The hash algorithm may be a secure hash algorithm (SHA) and a message digest algorithm version 5 (MD5), etc., which is not limited in this embodiment.

[0131] In one example, MD5 (message digest algorithm) may be performed on the device information or the node information, and 4 bytes of the MD5 value from low to high may be used as the first hash value of the device information or the second hash value of the node information.

[0132] In a specific embodiment, a hash operation is performed on the node information of the machine node, and the hash result (i.e., the second hash value) is divided by 2^32 and the remainder is taken. The remainder obtained must be an integer between 0 and 2^32-1. Therefore, there must be a point on the hash ring corresponding to this integer. This integer can be used to represent the machine node, that is, the machine node can be mapped to this hash ring.

[0133] For the candidate push device, its device information can be used as the key, and the device information is hashed using the same hash algorithm. The hash result (i.e., the first hash value) is divided by 2^32 and the remainder is taken. The remainder represents the candidate push device, and based on the remainder, the candidate push device is mapped to the hash ring.

[0134] Optionally, in this embodiment, the step of “searching for a target machine node for processing the candidate push device from each machine node according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space” may include:

[0135] Starting from the position corresponding to the device mapping point, searching for nodes along the hash space ring in a preset direction;

[0136] When a node mapping point is found, the machine node corresponding to the node mapping point is determined as a target machine node for processing the candidate push device.

[0137] The preset direction may be clockwise or counterclockwise, and may be set according to actual conditions.

[0138] Specifically, starting from the position of the device mapping point, the machine node corresponding to the first node mapping point in the clockwise direction is the target machine node for processing the candidate push device.

[0139] In this embodiment, the consistent hashing algorithm can be used to organize the entire hash value space into a virtual ring in a clockwise direction, called a hash (space) ring. The hash value space can be a 2^32 numerical space, and then the hash function is used to perform a hash operation on each machine node. Specifically, the IP or host name of the machine node can be selected as a keyword for hash operation, so as to determine the position of each machine node on the hash ring; finally, the same hash function is used to perform a hash operation on the candidate push device (key), and the position of this candidate push device on the hash ring is determined. From this position, the first machine node encountered in a clockwise direction along the hash ring is the machine node to be located.

[0140] The consistent hashing algorithm can solve the data hashing problem caused by adding or reducing machine nodes by reducing the impact range, thereby solving the load balancing problem in a distributed environment. If there is hot data, the hot area can be divided by adding nodes, and the pressure can be distributed to other machine nodes to achieve load balancing again.

[0141] Optionally, in this embodiment, the content push method may further include:

[0142] When an abnormality occurs in the target machine node, obtaining at least one backup machine node of the target machine node;

[0143] Based on the number of the standby machine nodes, the candidate push devices assigned to the target machine node are divided to obtain at least one device subset, wherein the device subset includes at least one candidate push device;

[0144] The at least one device subset is respectively allocated to each spare machine node, and each spare machine node is determined as a new target machine node for processing the allocated device subset.

[0145] Specifically, when the target machine node is abnormal, the traffic allocated to the target machine node will be redistributed to its corresponding standby machine node. The redistribution method can be an average distribution, etc., which is not limited in this embodiment. Specifically, the standby machine node can obtain the historical push data of the candidate push device allocated to it from the database corresponding to the target machine node.

[0146] Optionally, in this embodiment, before the step of “assigning the candidate push device to the target machine node”, the following may also be included:

[0147] Determining, according to the content attribute information of the target content, the sensitivity of the target content to push delay;

[0148] Based on the sensitivity, determining the priority information of the target content, and adding the push task corresponding to the target content to the message queue corresponding to the priority information;

[0149] Taking out the push task of the target content from the message queue corresponding to the priority information;

[0150] The step of “assigning the candidate push device to the target machine node” may include:

[0151] Based on the push task of the target content, the candidate push device is assigned to the target machine node.

[0152] The content attribute information of the target content may include its content type, such as whether it is news information, etc. The sensitivity of the content with different content attribute information to push delay is different. For example, news information is more sensitive to push delay, while historical articles are less sensitive to push delay. The priority information corresponding to content with different sensitivities is also different. The priority of high-sensitivity content can be higher, while the priority of low-sensitivity content can be lower.

[0153] 104. Performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine a target push device from the assigned candidate push devices.

[0154] In a specific content push scenario, in order to improve user experience, it is necessary to minimize the amount of content that users are not interested in receiving. Therefore, for each candidate push device assigned to the target machine node, a business demand analysis is required to determine the target push device that really needs to be pushed from these candidate push devices.

[0155] Specifically, business demand analysis may include push duplicate filtering, push frequency control, and dispersion.

[0156] Among them, push duplicate filtering is to prevent users from receiving duplicate push content. As various users, they hope to see interesting and fresh push content; as service providers of various push content applications, they hope to push content that users are interested in to users as much as possible and try not to let users receive the same push content.

[0157] Among them, push frequency control is to prevent users from receiving content pushes too frequently and to prevent the interval between two adjacent pushes received by users from being too short.

[0158] The purpose of breaking up is to prevent users from continuously receiving the same type of content push notifications and to ensure the diversity of push notifications received by users.

[0159] Optionally, in this embodiment, the step of “performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine the target push device from the assigned candidate push devices” may include:

[0160] For each candidate push device assigned to the target machine node, determine a push record logic block of the candidate push device, wherein the push record logic block includes at least one bit, and state identification information of the bit represents interaction information corresponding to the candidate push device and the content;

[0161] Performing a hash operation on the target content to obtain a content hash value;

[0162] Mapping the target content to the push record logic block based on the content hash value, and determining a mapping position of the target content in the push record logic block;

[0163] The candidate push devices are filtered according to the status identification information of the bit corresponding to the mapping position to obtain the target push device.

[0164] The push record logic block may be a Bloom block, which may record the push data of the candidate push device, and specifically record whether a certain content has been pushed to the candidate push device. In this embodiment, push duplication filtering may be performed through a Bloom filter. A Bloom filter is specifically a very long binary vector and a series of random mapping functions, wherein the binary vector may be regarded as a Bloom block, and the mapping function may be used to map the target content to the Bloom block.

[0165] The bit status identification information may include two types: set and unset. Specifically, the bit status identification information may be 1, and the bit status identification information may be 0.

[0166] The interaction information may indicate whether a certain content has been pushed to the candidate push device.

[0167] In some embodiments, several hash algorithms may be used to perform hash operations on the target content to obtain a content hash value, and then the content hash value is mapped to several bits of the push record logic block. If these bits have been set, it means that the target content has been pushed to the candidate push device before. Therefore, the candidate push device can be filtered out and the target content does not need to be pushed to the candidate push device this time.

[0168] Optionally, in this embodiment, the step of “filtering the candidate push devices according to the state identification information of the bit corresponding to the mapping position to obtain the target push device” may include:

[0169] When the status identification information of each bit corresponding to the mapping position indicates that each bit has been set, filtering out the candidate push device;

[0170] When the status identification information of each bit corresponding to the mapping position indicates that there is an unset bit among the bits, the candidate push device is retained, and the status identification information of the bit is updated to be set, so as to obtain the target push device corresponding to the target machine node.

[0171] The bit status identification information is updated to be set, specifically, the bit is changed from 0 to 1. The target push device corresponding to the target machine node is the retained candidate push device.

[0172] Optionally, in this embodiment, the step of “performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine the target push device from the assigned candidate push devices” may include:

[0173] For each candidate push device assigned to the target machine node, obtaining historical push data of the candidate push device in at least one dimension through the target machine node;

[0174] When the historical push data in the at least one dimension meets a preset condition, the candidate push device is determined as a target push device.

[0175] The historical push data of the candidate push device in at least one dimension may include historical push time, frequency information, push content type, etc., which is not limited in this embodiment.

[0176] The preset conditions can be set according to the actual situation. For example, it can be set that the last push time is more than 2 hours from the current time, or that the content push frequency information within 24 hours is not more than 8 times, or that the content push frequency information of a certain type within a preset time period is not more than 5 articles, etc.

[0177] The step of “determining the candidate push device as the target push device when the historical push data in at least one dimension meets a preset condition” may include:

[0178] Determining, based on historical push data, a predicted probability that the candidate push device triggers the target content;

[0179] When the predicted probability meets a preset condition, the candidate push device is determined as a target push device.

[0180] The predicted probability may reflect the probability that the user corresponding to the candidate push device clicks on the target content. Specifically, when the predicted probability exceeds a preset threshold, the candidate push device may be determined as the target push device.

[0181] 105. Push the target content to the target push device.

[0182] In a specific business scenario, the application can divide the push delivery link into triggering, recalling, scheduling, push delivery and other processes, and is also divided into several subsystems in terms of architecture. Specifically, the content push system of the application can include four subsystems, namely, a push trigger subsystem, a push recall subsystem, a push scheduling subsystem, and a push delivery subsystem. Figure 1c As shown, the corresponding system framework diagram is shown.

[0183] Among them, the push trigger subsystem is mainly developed according to specific business logic. In actual business, there are various push requirements, such as push triggers for user interaction (such as attention / comment), push triggers for identifying new users, interest automation triggers based on object interests, and triggers of push operation platforms. The push trigger process of the push trigger subsystem can be customized according to actual business needs. Push recall is an important part of implementing push, and the speed of recall determines the speed of push; for a piece of news to be pushed, after the news operation / recommendation algorithm selects the crowd tag to be pushed (i.e., the push object group identifier in the above embodiment), the push recall subsystem can quickly obtain the user device list corresponding to the crowd tag. After the business triggers a push, the push scheduling subsystem can determine whether the push content can be sent to the user (specifically, the user device list recalled by the push recall subsystem) to ensure user experience; specifically, the push scheduling subsystem can encapsulate some general judgment logic such as filtering, frequency control, avoidance, and dispersion. For push notifications that can be sent based on push scheduling decisions, the push notification subsystem can be used to connect to the device manufacturer's channel. The push message passes through the manufacturer's server and then reaches the user through the manufacturer's server.

[0184] The method for rapid recall of push users designed in this application can support news to quickly find matching user groups. Specifically, by cutting the crowd package into small file segments, the mapping relationship between the object group identifier (such as interest tags, mobile phone brands) and the crowd package file segment (specifically, the above-mentioned device set) can be established in the memory, and high-concurrency crowd package loading can be achieved through multi-process / multi-coroutine, and rapid recall and rapid delivery of crowd packages can be achieved; in addition, in order to improve the push and delivery speed, node local and other cache processes are also established; based on consistent hashing and priority queues, global and reasonable scheduling of news push messages is achieved to ensure hot spot bursts, etc., while achieving load balancing and high throughput of nodes within the cluster.

[0185] refer to Figure 1c , the four subsystems of the content push system can be specifically described as follows:

[0186] The push trigger subsystem can be responsible for triggering different business scenarios. In actual business scenarios, there are content pushes actively triggered by the push operation platform, special trigger processes for identifying new users, automated trigger processes for automatically pushing content to users based on user interests, and content pushes generated by interactive operations such as user attention / comments. In these business scenarios, when content is pushed, you only need to call the push delivery interface to push it. The push trigger subsystem only needs to pay attention to the business trigger logic and specify the audience tag of the content to be pushed; the specific logic processing of the push can be handled by other subsystems. Among them, audience tags can be mainly divided into two categories, one is interest tags, and the other is the geographical tag of the object.

[0187] The push recall subsystem can receive the content to be pushed and the tags of the audience users to be pushed specified by the upstream push trigger subsystem; the push recall subsystem can decompose the audience tags and convert them into a specific user device list, which depends on the user mining process.

[0188] Specifically, the user's interactive operations on the content in the content distribution APP (such as browsing / clicking / commenting, etc.) can be analyzed to determine the content that the user is more interested in, and the user's interactive attribute information can be obtained. These interactive attribute information will be recorded in the user's object attribute base table. This is a streaming process. As the user continues to use the content distribution APP, the interactive operations generated continuously update the object attribute base table. In some embodiments, a timed process can be started to scan the object attribute base table at dawn every day, and a user device list (i.e., the original device set in the above embodiment) is aggregated one by one according to the user's interactive attribute information, and these device lists are written into number package files one by one, and these device lists are uploaded to the distributed file system. In order to facilitate filtering and parallel processing of these device lists, these device lists can be segmented as follows:

[0189] a) For the device list under the interest tag type (which can be regarded as an interest user package), it can be divided into multiple small device sets according to the type of device used by the user (such as mobile phone brand). Specifically, each interest tag corresponds to a directory, and each directory is divided into several mobile phone brand sub-directories according to the user's mobile phone brand. Separation by brand is convenient for quick screening of user packages based on mobile phone brands; each brand sub-directory is a specific user device list, and each specific user device list can be cut into several small files (device sets) according to 15,000 devices. Among them, the small file segmentation is to facilitate parallel task processing.

[0190] like Figure 1d As shown, the device list under interest tags such as "Society" and "Finance" can be divided into multiple small device collections according to mobile phone brands.

[0191] b) For the device list under the regional tag type (which can be regarded as a regional user package), it can be divided into multiple small device sets according to the login address of the user's device. Specifically, the device list can be stored in two directories at the province / prefecture-level city level. Each prefecture-level city has a specific user device list file, which can also be divided into several small files according to 15,000 devices; therefore, the user device list supports direct query by province / prefecture-level city. If you want to query directly by county / district, you need to filter in memory based on the prefecture-level city.

[0192] like Figure 1d As shown, the device list under the "Province K" regional label can be divided into multiple small device sets by city.

[0193] In order to conveniently manage these number package files (i.e., device collections, or user device packages), a package management service can be built to manage the acquisition of device collections. The services provided by the package management service can be as follows:

[0194] a) Data synchronization: The package management service can periodically check the version of the user device package on the remote distributed file system. If the version of the remote device package is newer than the local version, data synchronization will be initiated to synchronize the user device package on the remote distributed file system to the local version.

[0195] b) The package management service also stores metadata information such as the organizational structure, data version, data update time, total number of specific number package files, and path information of the user device package in the memory. When retrieving the user device package, the path where the optimized device package is located can be located from the metadata information recorded in the memory, and the corresponding user device package content can be directly read from the disk and returned;

[0196] c) The package management service can deploy multiple node copies, each node can cache a set of devices with preset object group identifiers, and each service node can report heartbeats to the name discovery service. The name discovery service will automatically remove faulty service nodes to prevent upstream services from accessing faulty nodes. When the upstream obtains a specific number package fragment, if a node reads the number package fragment abnormally, it will report an error to the upstream, and the upstream will retry to other available nodes. This embodiment can ensure the high availability of the package management service through two methods: "name discovery service automatically removes abnormal nodes" and "upstream request fails and retry", and the service availability rate is above 99.99%.

[0197] In addition, because multiple node copies are introduced, the consistency of data versions of multiple copies needs to be maintained. This embodiment can introduce a version detection sentinel service. The sentinel service can detect the data version of each package management service. When it is found that the version of the service node in the cluster is inconsistent, the sentinel service sends a data synchronization signal to all service nodes, and all service nodes initiate data synchronization to the remote server. This can refresh all service nodes in the cluster to the latest data version. When the service node fails to synchronize data, it means that there may be a problem with this node. At this time, it will report its abnormal status to the name discovery service, and then the name discovery service will remove the node to prevent upstream services from accessing the abnormal node.

[0198] The push scheduling subsystem is mainly composed of a scheduling access module and a scheduling service module.

[0199] Among them, the scheduling access module can be used to access push messages. Specifically, it can accept message tasks from the upstream push trigger subsystem and send them to the scheduling service module for cluster task processing; different push content has different sensitivities to push delays, so when processing push messages, a priority queue can be used here. Hot and sudden content pushes enjoy high push priority and can be scheduled and processed first.

[0200] In addition, the scheduling access module can also be used to perform cluster task allocation for the scheduling service module. Specifically, after taking out the push task from the priority queue, it needs to be sent to the specific scheduling service node (i.e., the machine node in the above embodiment) for scheduling; in order to improve the push delivery speed, the scheduling service module caches the object attribute data (specifically, it can be historical push data, such as the user's historical push content, the user's push preference settings on the client, etc.) in the process local (i.e., each machine node local); when making push task allocation, in order to avoid local cache failure, a consistent hashing algorithm is used, and the push task of the same user will be mapped to a fixed machine node, but when the machine node responds to timeout, it means that the machine node is overloaded or the service is abnormal, and the push task will be sent to the standby node of the machine node. Each machine node can have 4 fixed standby nodes. When the machine node is abnormal, the traffic to the machine node will be evenly distributed to the 4 standby nodes to achieve failover.

[0201] Among them, the scheduling service module mainly executes a specific scheduling process for the push tasks distributed by the scheduling access module. The scheduling process may include push duplicate filtering, push frequency control, quota management, dispersion, channel management and protocol customization, as described below.

[0202] Among them, push duplicate filtering is to prevent users from receiving duplicate push content. Specifically, each news to be pushed has a news identification number, and the news identification number can be used as a push filtering key to ensure that the same news identification number will not be pushed to the user twice. In the specific implementation of duplicate filtering, Bloom filtering can be used. Each user (or device) corresponds to a Bloom block. Through several hash algorithms, the news identification number of the article is mapped to several bits of the Bloom block. If these bits are set, it means that the article has been sent to the user, and this push will be filtered out; otherwise, it means that the article has not been pushed to the user before, and the other processes below continue, and these bits are set at the same time.

[0203] Push frequency control can prevent users from receiving push notifications too frequently and prevent the interval between two adjacent push notifications from being too short.

[0204] Quota management can specifically manage the total number of pushes. This is because different device manufacturers have different requirements for the number and frequency of news pushes. For example, some manufacturers limit the total number of pushes per day, while some manufacturers limit the total number of pushes per user per day. Quota management can be based on the total number of daily push quotas, and pushes can be dispersed into different time periods throughout the day. By predicting the likelihood of users clicking on pushes, pushes with low conversion rates can be filtered out to ensure that the quota is not quickly used up due to intensive pushes in a certain time period, while improving the conversion rate of pushes.

[0205] Breaking up the push notifications can prevent users from receiving the same type of push notifications continuously, ensuring that the push notifications received by users are diverse and surprising.

[0206] Channel management and protocol customization specifically refer to selecting the specific channel for push messages that are confirmed and sent after scheduling, as well as customizing the push protocol.

[0207] Among them, the push task can be given to the push sending subsystem after scheduling; the push sending subsystem can finally reach the user through the long link channel and the manufacturer channel corresponding to the user's mobile phone brand.

[0208] Among them, the long link channel means that after the link connection is established, the link is not actively disconnected when the link is idle, but the link connection is always maintained.

[0209] The present application proposes an architectural design method for a news push system that can achieve rapid distribution, which can decouple a complex push process into four subsystems: push triggering, push recall, push scheduling, and push distribution. The subsystems have high cohesion and low coupling between subsystems. Various push requirements in different scenarios only need to implement customized push triggering according to business requirements. When the push is triggered, only the name of the object group to be sent needs to be specified, and there is no need to worry about how to find the audience objects of the push. The push recall process is responsible for replacing the object group name with a specific audience device list. There is no need to worry about the implementation of complex rules such as filtering duplicates, frequency control, avoidance, and dispersion. These rules can be encapsulated in a unified push scheduling module. The push scheduling is responsible for accessing the push triggered by various scenarios and performing unified push scheduling based on users. If the push meets the various rule conditions such as frequency control, avoidance, and dispersion, it can enter the push distribution process. The push distribution process can push content to users.

[0210] Specifically, users obtain news information mainly through two methods: pull and push. In the former, users actively log in to the news app to check the content flow (feed flow); in the latter, when there is a hot topic or news that users are interested in, push messages are actively sent to users who may be interested. News push is an important means for users to obtain news information in a timely manner. Because users cannot frequently check the news app to obtain the latest news and information, compared with users actively querying news information, news push helps users obtain the latest news and information more timely, so news push is one of the most important ways of news distribution.

[0211] For news push, it is more important to reflect "newness". Through questionnaire surveys on users, users are most concerned about the timeliness of "push messages" for news push. Users may click on push messages that they are interested in for the first time, and they will not click on push messages that they have seen before; therefore, the delivery speed is also an important optimization goal for news push. The content push method of this application focuses on how to improve the delivery speed of news push while ensuring user experience. Through comparative experiments, it is found that based on the content push method of this application, the time consumption of the news delivery link is optimized by 50%, and the number of clicks on news push can be increased by about 10%.

[0212] For the push of a piece of content, this application can select audience users with matching interest tags through the push recall process to improve the conversion rate of the push; in addition, when doing message push, this application can implement calculations such as filtering duplicates, controlling frequency, and breaking up through the push scheduling process, thus ensuring the user experience. For example, it can prevent users from receiving repeated content pushes, and by controlling the push frequency to avoid frequently disturbing users, it can also ensure the diversity of push messages received by users to ensure the surprise of users receiving pushes.

[0213] As can be seen from the above, this embodiment can determine the target content to be pushed, and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; based on the device set corresponding to the push object group identifier and the preset object group identifier, recall at least one push device set corresponding to the target content, and the push device set includes at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; through the target machine node, perform business demand analysis on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and push the target content to the target push device. When triggering the push of target content, the present application can determine the corresponding push device set based on the push object group identifier specified by the target content, without having to find the push audience object based on the object attribute information of massive candidate objects during the content push process, which greatly improves the speed of determining the candidate push devices; in addition, the candidate push devices can be quickly allocated and sent to the corresponding machine nodes based on the association relationship between the mapping points of the candidate push devices and the machine nodes in the preset data space, thereby achieving load balancing of each machine node, and thus performing business demand analysis on the allocated candidate push devices through the machine nodes, so as to push content to the devices that really need to be pushed, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy.

[0214] According to the method described in the above embodiment, the following will be further described in detail by taking the example of the content push device being specifically integrated in the server.

[0215] The present application embodiment provides a content push method, such as Figure 2 As shown, the specific process of the content push method can be as follows:

[0216] 201. The server determines target content to be pushed and a push object group identifier corresponding to the target content; and determines at least one machine node for pushing the target content.

[0217] The push target group identifier may be the tag information of the recommended target group, specifically the audience tag of the content to be pushed. For example, the push target group identifier may be the target interest, the device used (such as the type of mobile phone), the region where the target is located, etc., which is not limited in this embodiment.

[0218] Among them, the machine node can be a server node used to schedule push tasks. Specifically, it can be used to perform business demand analysis on each assigned push task to determine the tasks that really need to be pushed. The business demand analysis may include push duplicate filtering, push frequency control, quota management, and dispersion.

[0219] 202. The server recalls at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier, where the push device set includes at least one candidate push device.

[0220] Among them, the matching degree between the push object group identifier and each preset object group identifier can be determined, and the device set of the preset object group identifier whose matching degree meets the preset matching conditions is determined as the device set corresponding to the push object group identifier. The device set corresponding to the push object group identifier is also the push device set corresponding to the target content.

[0221] The candidate push device may be a device used by the candidate push user, such as a mobile phone, a computer, etc.

[0222] Optionally, in this embodiment, the content push method may further include:

[0223] Acquiring interactive data, wherein the interactive data includes interactive attribute information of at least one interactive object;

[0224] Based on the interactive attribute information of each interactive object, aggregate the devices using the interactive object to obtain at least one original device set;

[0225] For each original device set, based on at least one preset object group identifier, the original device set is segmented to obtain device sets corresponding to each preset object group identifier.

[0226] The interactive object may be an object that interacts with the content, and the interaction here may include browsing, clicking, commenting, etc. The interactive attribute information may include preference information of the interactive object, etc.

[0227] Among them, the step of "based on the interactive attribute information of each interactive object, aggregating the usage devices of the interactive object to obtain at least one original device set" can specifically be: according to the interactive attribute information, aggregating each interactive object to obtain multiple object sets, each object set includes at least one interactive object, and different object sets may correspond to different types of interactive objects, where the type may refer to the type of interest preference, etc.; then, based on the usage devices of each interactive object, the interactive objects in the object set are replaced with their corresponding usage devices to obtain the corresponding original device set.

[0228] After obtaining the original device set, since the original device set may still contain a large amount of data, it can be cut into several small device sets according to each preset object group identifier. Segmentation of the original device set can also facilitate task parallel processing.

[0229] This embodiment can decompose and convert the push object group identifier into a specific user device list (ie, a push device set), which can greatly increase the speed of determining the push object, thereby improving the content push efficiency.

[0230] 203. The server maps each candidate push device to a preset data space based on the device information of the candidate push device, and obtains a device mapping point of the candidate push device in the preset data space.

[0231] The device information of the candidate push device may be a device name or the like.

[0232] 204. The server maps each machine node to the preset data space based on the node information of each machine node, and obtains a node mapping point of each machine node in the preset data space.

[0233] The node information of the machine node may be the IP (Internet Protocol Address) or host name of the machine node.

[0234] Optionally, in this embodiment, the preset data space is a hash space ring composed of hash values;

[0235] The step of “mapping the candidate push device to a preset data space based on the device information of the candidate push device to obtain a device mapping point of the candidate push device in the preset data space” may include:

[0236] Performing a hash operation on the device information of the candidate push device to obtain a first hash value;

[0237] According to the first hash value, mapping the candidate push device to the hash space ring to obtain a device mapping point of the candidate push device on the hash space ring;

[0238] The step of “mapping each machine node to the preset data space based on the node information of each machine node to obtain a node mapping point of each machine node in the preset data space” may include:

[0239] Performing a hash operation on the node information of each machine node to obtain a second hash value;

[0240] According to the second hash value, each machine node is mapped to the hash space ring to obtain a node mapping point of each machine node on the hash space ring.

[0241] Specifically, the hash space ring can be a circular ring corresponding to the hash space of 2^32-1, that is, it is a circle composed of 2^32 points, and each point on the circle represents 1, 2, 3, ... until 2^32-1.

[0242] The hash algorithm used to perform hash operations on the device information of the candidate push device and the node information of the machine node may be the same. The hash algorithm may be a secure hash algorithm (SHA) and a message digest algorithm version 5 (MD5), etc., which is not limited in this embodiment.

[0243] In a specific embodiment, a hash operation is performed on the node information of the machine node, and the hash result (i.e., the second hash value) is divided by 2^32 and the remainder is taken. The remainder obtained must be an integer between 0 and 2^32-1. Therefore, there must be a point on the hash ring corresponding to this integer. This integer can be used to represent the machine node, that is, the machine node can be mapped to this hash ring.

[0244] For the candidate push device, its device information can be used as the key, and the device information is hashed using the same hash algorithm. The hash result (i.e., the first hash value) is divided by 2^32 and the remainder is taken. The remainder represents the candidate push device, and based on the remainder, the candidate push device is mapped to the hash ring.

[0245] 205. The server searches for a target machine node for processing the candidate push device from the machine nodes according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space, and assigns the candidate push device to the target machine node.

[0246] Optionally, in this embodiment, the step of “searching for a target machine node for processing the candidate push device from each machine node according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space” may include:

[0247] Starting from the position corresponding to the device mapping point, searching for nodes along the hash space ring in a preset direction;

[0248] When a node mapping point is found, the machine node corresponding to the node mapping point is determined as a target machine node for processing the candidate push device.

[0249] The preset direction may be clockwise or counterclockwise, and may be set according to actual conditions.

[0250] Specifically, starting from the position of the device mapping point, the machine node corresponding to the first node mapping point in the clockwise direction is the target machine node for processing the candidate push device.

[0251] In this embodiment, the consistent hashing algorithm can be used to organize the entire hash value space into a virtual ring in a clockwise direction, called a hash (space) ring. The hash value space can be a 2^32 numerical space, and then the hash function is used to perform a hash operation on each machine node. Specifically, the IP or host name of the machine node can be selected as a keyword for hash operation, so as to determine the position of each machine node on the hash ring; finally, the same hash function is used to perform a hash operation on the candidate push device (key), and the position of this candidate push device on the hash ring is determined. From this position, the first machine node encountered in a clockwise direction along the hash ring is the machine node to be located.

[0252] The consistent hashing algorithm can solve the data hashing problem caused by adding or reducing machine nodes by reducing the impact range, thereby solving the load balancing problem in a distributed environment. If there is hot data, the hot area can be divided by adding nodes, and the pressure can be distributed to other machine nodes to achieve load balancing again.

[0253] 206. The server performs a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node, so as to determine a target push device from the assigned candidate push devices.

[0254] Specifically, business demand analysis may include push duplicate filtering, push frequency control, and dispersion.

[0255] Optionally, in this embodiment, the step of “performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine the target push device from the assigned candidate push devices” may include:

[0256] For each candidate push device assigned to the target machine node, determine a push record logic block of the candidate push device, wherein the push record logic block includes at least one bit, and state identification information of the bit represents interaction information corresponding to the candidate push device and the content;

[0257] Performing a hash operation on the target content to obtain a content hash value;

[0258] Mapping the target content to the push record logic block based on the content hash value, and determining a mapping position of the target content in the push record logic block;

[0259] The candidate push devices are filtered according to the status identification information of the bit corresponding to the mapping position to obtain the target push device.

[0260] The push record logic block may be a Bloom block, which may record the push data of the candidate push device, and specifically record whether a certain content has been pushed to the candidate push device. In this embodiment, push duplication filtering may be performed through a Bloom filter. A Bloom filter is specifically a very long binary vector and a series of random mapping functions, wherein the binary vector may be regarded as a Bloom block, and the mapping function may be used to map the target content to the Bloom block.

[0261] The bit status identification information may include two types: set and unset. Specifically, the bit status identification information may be 1, and the bit status identification information may be 0.

[0262] The interaction information may indicate whether a certain content has been pushed to the candidate push device.

[0263] In some embodiments, several hash algorithms may be used to perform hash operations on the target content to obtain a content hash value, and then the content hash value is mapped to several bits of the push record logic block. If these bits have been set, it means that the target content has been pushed to the candidate push device before. Therefore, the candidate push device can be filtered out and the target content does not need to be pushed to the candidate push device this time.

[0264] Optionally, in this embodiment, the step of “filtering the candidate push devices according to the state identification information of the bit corresponding to the mapping position to obtain the target push device” may include:

[0265] When the status identification information of each bit corresponding to the mapping position indicates that each bit has been set, filtering out the candidate push device;

[0266] When the status identification information of each bit corresponding to the mapping position indicates that there is an unset bit among the bits, the candidate push device is retained, and the status identification information of the bit is updated to be set, so as to obtain the target push device corresponding to the target machine node.

[0267] The bit status identification information is updated to be set, specifically, the bit is changed from 0 to 1. The target push device corresponding to the target machine node is the retained candidate push device.

[0268] 207. The server pushes the target content to the target push device.

[0269] The content push method provided in this application is mainly aimed at some content pushes with high "timeliness", such as news push. News push is different from ordinary personalized content push. The biggest feature of news is "newness", and news push must also reflect timeliness. Therefore, the most important requirement for news push is to achieve rapid delivery. This application uses the architectural design of a news push system with rapid delivery as the goal, and can achieve "first push on the entire network" for the push of many hot emergencies. Specifically, this application realizes a fast delivery link for news push by optimizing the architectural design of multiple module processes such as push recall and push scheduling. For a large hot emergency push to all active users, the content push system provided by this application can complete the push within 1 minute.

[0270] As can be seen from the above, in this embodiment, the target content to be pushed and the push object group identifier corresponding to the target content can be determined by the server; and at least one machine node for pushing the target content can be determined; based on the push object group identifier and the device set corresponding to the preset object group identifier, at least one push device set corresponding to the target content is recalled, and the push device set includes at least one candidate push device; for each candidate push device, based on the device information of the candidate push device, the candidate push device is mapped to the preset data space to obtain the device mapping point of the candidate push device in the preset data space; based on the node information of each machine node, each machine node is mapped to the preset data space to obtain the node mapping point of each machine node in the preset data space; according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space, a target machine node for processing the candidate push device is searched from each machine node, and the candidate push device is assigned to the target machine node; through the target machine node, a business demand analysis is performed on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and the target content is pushed to the target push device.

[0271] When triggering the push of target content, the present application can determine the corresponding push device set based on the push object group identifier specified by the target content, without having to find the push audience object based on the object attribute information of massive candidate objects during the content push process, which greatly improves the speed of determining the candidate push devices; in addition, the candidate push devices can be quickly allocated and sent to the corresponding machine nodes based on the association relationship between the mapping points of the candidate push devices and the machine nodes in the preset data space, thereby achieving load balancing of each machine node, and thus performing business demand analysis on the allocated candidate push devices through the machine nodes, so as to push content to the devices that really need to be pushed, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy.

[0272] In order to better implement the above method, the embodiment of the present application also provides a content push device, such as Figure 3 As shown, the content push device may include a determination unit 301, a recall unit 302, a node determination unit 303, an analysis unit 304, and a push unit 305, as follows:

[0273] (1) determining unit 301;

[0274] The determination unit is used to determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content.

[0275] (2) recall unit 302;

[0276] The recall unit is used to recall at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier, wherein the push device set includes at least one candidate push device.

[0277] Optionally, in some embodiments of the present application, the content push apparatus may further include a device set determination unit as follows:

[0278] The device set determination unit is used to obtain interaction data, where the interaction data includes interaction attribute information of at least one interactive object; based on the interaction attribute information of each interactive object, the devices using the interactive object are aggregated to obtain at least one original device set; for each original device set, based on at least one preset object group identifier, the original device set is segmented to obtain device sets corresponding to each preset object group identifier.

[0279] Optionally, in some embodiments of the present application, the device set corresponding to the preset object group identifier is stored in a remote server;

[0280] The content push device may further include a cache unit, as follows:

[0281] The cache unit is used to cache the device set corresponding to the preset object group identifier stored in the remote server to the preset service storage cluster, so that at least two service nodes in the preset service storage cluster cache the device set corresponding to the preset object group identifier;

[0282] The recall unit may include a reading subunit, a re-reading subunit and a determining subunit, as follows:

[0283] The reading subunit is used to read the device set corresponding to the push object group identifier through a target service node, and the target service node is any service node;

[0284] A re-reading subunit is used for, when reading is abnormal, performing an identification query in the reference service node based on the push object group identifier through a reference service node to read the device set corresponding to the push object group identifier, wherein the reference service node is a service node other than the target service node;

[0285] The determination subunit is configured to determine the read device set as at least one push device set corresponding to the target content.

[0286] Optionally, in some embodiments of the present application, the content push device may further include a version detection unit, as follows:

[0287] The version detection unit is used to detect the version information of the device set stored in each service node; when it is detected that the version information of the device set in the service node is inconsistent, the service node is triggered to send a data synchronization instruction to the remote server; based on the data synchronization instruction, the device set currently stored in the remote server is synchronized to each service node.

[0288] (3) Node determination unit 303;

[0289] The node determination unit is used to determine, for each candidate push device, a target machine node for processing the candidate push device from among the machine nodes based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node in a preset data space, and assign the candidate push device to the target machine node.

[0290] Optionally, in some embodiments of the present application, the node determination unit may include a first mapping subunit, a second mapping subunit and a node determination subunit, as follows:

[0291] The first mapping subunit is used to map the candidate push device to a preset data space based on the device information of the candidate push device, so as to obtain a device mapping point of the candidate push device in the preset data space;

[0292] A second mapping subunit is used to map each machine node to the preset data space based on the node information of each machine node, so as to obtain a node mapping point of each machine node in the preset data space;

[0293] The node determination subunit is used to search for a target machine node for processing the candidate push device from the machine nodes according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space.

[0294] Optionally, in some embodiments of the present application, the preset data space is a hash space ring composed of hash values;

[0295] The first mapping subunit may be specifically configured to perform a hash operation on the device information of the candidate push device to obtain a first hash value; map the candidate push device to the hash space ring according to the first hash value to obtain a device mapping point of the candidate push device on the hash space ring;

[0296] The second mapping subunit can be specifically used to perform hash operation on the node information of each machine node to obtain a second hash value; according to the second hash value, map each machine node to the hash space ring to obtain a node mapping point of each machine node on the hash space ring.

[0297] Optionally, in some embodiments of the present application, the node determination subunit can be specifically used to start from the position corresponding to the device mapping point, and perform node search along the hash space ring in a preset direction; when the node mapping point is found, the machine node corresponding to the node mapping point is determined as the target machine node for processing the candidate push device.

[0298] Optionally, in some embodiments of the present application, the content push device may further include a standby node replacement unit, as follows:

[0299] The standby node replacement unit is configured to obtain at least one standby machine node of the target machine node when an abnormality occurs in the target machine node; divide the candidate push devices allocated to the target machine node based on the number of the standby machine nodes to obtain at least one device subset, wherein the device subset includes at least one candidate push device; allocate the at least one device subset to each standby machine node, and determine each standby machine node as a new target machine node for processing the allocated device subset.

[0300] Optionally, in some embodiments of the present application, the content push device may further include a priority unit as follows:

[0301] The priority unit is used to determine the sensitivity of the target content to the push delay according to the content attribute information of the target content; determine the priority information of the target content based on the sensitivity, add the push task corresponding to the target content to the message queue corresponding to the priority information; and take out the push task of the target content from the message queue corresponding to the priority information;

[0302] The node determination unit may be specifically configured to allocate the candidate push device to the target machine node based on the push task of the target content.

[0303] (4) analysis unit 304;

[0304] The analyzing unit is used to perform a business demand analysis on the candidate push devices allocated to the target machine node through the target machine node, so as to determine the target push device from the allocated candidate push devices.

[0305] Optionally, in some embodiments of the present application, the analysis unit may include a logic block determination subunit, a hash subunit, a third mapping subunit, and a filtering subunit, as follows:

[0306] The logic block determination subunit is used to determine, for each candidate push device assigned to the target machine node, a push record logic block of the candidate push device, wherein the push record logic block includes at least one bit, and state identification information of the bit represents interaction information corresponding to the candidate push device and the content;

[0307] A hash subunit, used to perform a hash operation on the target content to obtain a content hash value;

[0308] a third mapping subunit, configured to map the target content to the push record logic block based on the content hash value, and determine a mapping position of the target content in the push record logic block;

[0309] The filtering subunit is used to filter the candidate push devices according to the status identification information of the bit corresponding to the mapping position to obtain the target push device.

[0310] Optionally, in some embodiments of the present application, the filtering sub-unit can be specifically used to filter out the candidate push device when the status identification information of each bit corresponding to the mapping position indicates that each bit has been set; when the status identification information of each bit corresponding to the mapping position indicates that there is a bit that is not set among the bits, retain the candidate push device and update the status identification information of the bit to be set, so as to obtain the target push device corresponding to the target machine node.

[0311] Optionally, in some embodiments of the present application, the analysis unit may include a data acquisition subunit and a device determination subunit, as follows:

[0312] The data acquisition subunit is used to acquire, for each candidate push device assigned to the target machine node, historical push data of the candidate push device in at least one dimension through the target machine node;

[0313] The device determination subunit is used to determine the candidate push device as the target push device when the historical push data in the at least one dimension meets a preset condition.

[0314] (5) Push unit 305;

[0315] A pushing unit is used to push the target content to the target pushing device.

[0316] As can be seen from the above, in this embodiment, the determination unit 301 can determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; through the recall unit 302, based on the push object group identifier and the device set corresponding to the preset object group identifier, recall at least one push device set corresponding to the target content, and the push device set includes at least one candidate push device; through the node determination unit 303, for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; through the analysis unit 304, through the target machine node, perform business demand analysis on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and through the push unit 305, push the target content to the target push device. When triggering the push of target content, the present application can determine the corresponding push device set based on the push object group identifier specified by the target content, without having to find the push audience object based on the object attribute information of massive candidate objects during the content push process, which greatly improves the speed of determining the candidate push devices; in addition, the candidate push devices can be quickly allocated and sent to the corresponding machine nodes based on the association relationship between the mapping points of the candidate push devices and the machine nodes in the preset data space, thereby achieving load balancing of each machine node, and thus performing business demand analysis on the allocated candidate push devices through the machine nodes, so as to push content to the devices that really need to be pushed, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy.

[0317] The present application also provides an electronic device, such as Figure 4 As shown, it shows a schematic diagram of the structure of an electronic device involved in an embodiment of the present application, and the electronic device may be a terminal or a server, etc. Specifically:

[0318] The electronic device may include components such as a processor 401 with one or more processing cores, a memory 402 with one or more computer-readable storage media, a power supply 403, and an input unit 404. Those skilled in the art will appreciate that Figure 4 The electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0319] The processor 401 is the control center of the electronic device, and uses various interfaces and lines to connect various parts of the entire electronic device. By running or executing software programs and / or modules stored in the memory 402, and calling data stored in the memory 402, the processor 401 performs various functions of the electronic device and processes data. Optionally, the processor 401 may include one or more processing cores; preferably, the processor 401 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 401.

[0320] The memory 402 can be used to store software programs and modules. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 with access to the memory 402.

[0321] The electronic device also includes a power supply 403 for supplying power to each component. Preferably, the power supply 403 can be logically connected to the processor 401 through a power management system, so as to manage charging, discharging, power consumption and other functions through the power management system. The power supply 403 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators and other arbitrary components.

[0322] The electronic device may further include an input unit 404, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0323] Although not shown, the electronic device may further include a display unit, etc., which will not be described in detail herein. Specifically in this embodiment, the processor 401 in the electronic device will load the executable files corresponding to the processes of one or more application programs into the memory 402 according to the following instructions, and the processor 401 will run the application programs stored in the memory 402, thereby realizing various functions, as follows:

[0324] Determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; based on the device set corresponding to the push object group identifier and the preset object group identifier, recall at least one push device set corresponding to the target content, the push device set including at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; perform business demand analysis on the candidate push device assigned to the target machine node through the target machine node, so as to determine the target push device from the assigned candidate push devices; and push the target content to the target push device.

[0325] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0326] As can be seen from the above, this embodiment can determine the target content to be pushed, and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; based on the device set corresponding to the push object group identifier and the preset object group identifier, recall at least one push device set corresponding to the target content, and the push device set includes at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; through the target machine node, perform business demand analysis on the candidate push device assigned to the target machine node to determine the target push device from the assigned candidate push devices; and push the target content to the target push device. When triggering the push of target content, the present application can determine the corresponding push device set based on the push object group identifier specified by the target content, without having to find the push audience object based on the object attribute information of massive candidate objects during the content push process, which greatly improves the speed of determining the candidate push devices; in addition, the candidate push devices can be quickly allocated and sent to the corresponding machine nodes based on the association relationship between the mapping points of the candidate push devices and the machine nodes in the preset data space, thereby achieving load balancing of each machine node, and thus performing business demand analysis on the allocated candidate push devices through the machine nodes, so as to push content to the devices that really need to be pushed, which can greatly improve the speed and efficiency of content push while ensuring the push accuracy.

[0327] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0328] To this end, an embodiment of the present application provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the steps in any content push method provided in the embodiment of the present application. For example, the instructions can execute the following steps:

[0329] Determine the target content to be pushed and the push object group identifier corresponding to the target content; and determine at least one machine node for pushing the target content; based on the device set corresponding to the push object group identifier and the preset object group identifier, recall at least one push device set corresponding to the target content, the push device set including at least one candidate push device; for each candidate push device, based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine the target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; perform business demand analysis on the candidate push device assigned to the target machine node through the target machine node, so as to determine the target push device from the assigned candidate push devices; and push the target content to the target push device.

[0330] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0331] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0332] Since the instructions stored in the computer-readable storage medium can execute the steps in any content push method provided in the embodiments of the present application, the beneficial effects that can be achieved by any content push method provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0333] According to one aspect of the present application, a computer program product or a computer program is provided, the computer program product or the computer program including computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the methods provided in various optional implementations of the above-mentioned content push.

[0334] The above is a detailed introduction to a content push method and related equipment provided in an embodiment of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A content push method, characterized in that: include: Determine the target content to be pushed and the push object group identifier corresponding to the target content; and determining at least one machine node for pushing the target content; Based on the push object group identifier and the device set corresponding to the preset object group identifier, recall at least one push device set corresponding to the target content, the push device set including at least one candidate push device; For each candidate push device, based on the association between the device information of the candidate push device and the mapping points of the node information of each machine node on the preset data space, determine a target machine node for processing the candidate push device from the machine nodes, and assign the candidate push device to the target machine node; By means of the target machine node, a business demand analysis is performed on the candidate push devices assigned to the target machine node, so as to determine the target push device from the assigned candidate push devices; The target content is pushed to the target push device.

2. The method according to claim 1, characterized in that The method further comprises: Acquiring interactive data, wherein the interactive data includes interactive attribute information of at least one interactive object; Based on the interactive attribute information of each interactive object, aggregate the devices using the interactive object to obtain at least one original device set; For each original device set, based on at least one preset object group identifier, the original device set is segmented to obtain device sets corresponding to each preset object group identifier.

3. The method according to claim 1, characterized in that: The determining a target machine node for processing the candidate push device from the machine nodes based on the association relationship between the device information of the candidate push device and the mapping points of the node information of each machine node in the preset data space, comprises: Based on the device information of the candidate push device, the candidate push device is mapped to a preset data space to obtain a device mapping point of the candidate push device in the preset data space; Based on the node information of each machine node, each machine node is mapped to the preset data space to obtain a node mapping point of each machine node in the preset data space; According to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space, a target machine node for processing the candidate push device is searched from the machine nodes.

4. The method according to claim 3, characterized in that The preset data space is a hash space ring composed of hash values; The step of mapping the candidate push device to a preset data space based on the device information of the candidate push device to obtain a device mapping point of the candidate push device in the preset data space includes: Performing a hash operation on the device information of the candidate push device to obtain a first hash value; According to the first hash value, mapping the candidate push device to the hash space ring to obtain a device mapping point of the candidate push device on the hash space ring; The step of mapping each machine node to the preset data space based on the node information of each machine node to obtain a node mapping point of each machine node in the preset data space includes: Performing a hash operation on the node information of each machine node to obtain a second hash value; According to the second hash value, each machine node is mapped to the hash space ring to obtain a node mapping point of each machine node on the hash space ring.

5. The method according to claim 4, characterized in that The step of searching, from among the machine nodes, a target machine node for processing the candidate push device according to the positional relationship between the device mapping point and the node mapping point of each machine node in the preset data space comprises: Starting from the position corresponding to the device mapping point, searching for nodes along the hash space ring in a preset direction; When a node mapping point is found, the machine node corresponding to the node mapping point is determined as a target machine node for processing the candidate push device.

6. The method according to claim 1, characterized in that The method further comprises: When an abnormality occurs in the target machine node, obtaining at least one backup machine node of the target machine node; Based on the number of the standby machine nodes, the candidate push devices assigned to the target machine node are divided to obtain at least one device subset, wherein the device subset includes at least one candidate push device; The at least one device subset is respectively allocated to each spare machine node, and each spare machine node is determined as a new target machine node for processing the allocated device subset.

7. The method according to claim 1, characterized in that The device set corresponding to the preset object group identifier is stored in a remote server; Before recalling at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier, the method further includes: Caching the device set corresponding to the preset object group identifier stored in the remote server to the preset service storage cluster, so that at least two service nodes in the preset service storage cluster cache the device set corresponding to the preset object group identifier; The recalling at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier includes: Reading a device set corresponding to the push object group identifier through a target service node, wherein the target service node is any service node; When reading is abnormal, through a reference service node, based on the push object group identifier, an identifier query is performed in the reference service node to read the device set corresponding to the push object group identifier, and the reference service node is a service node other than the target service node; The read device set is determined as at least one push device set corresponding to the target content.

8. The method according to claim 7, characterized in that The method further comprises: Detect the version information of the device set stored in each service node; When it is detected that the version information of the device set in the service node is inconsistent, triggering the service node to send a data synchronization instruction to the remote server; Based on the data synchronization instruction, the device set currently stored in the remote server is synchronized to each service node.

9. The method according to claim 1, characterized in that: Before allocating the candidate push device to the target machine node, the method further includes: Determining, according to the content attribute information of the target content, the sensitivity of the target content to push delay; Based on the sensitivity, determining the priority information of the target content, and adding the push task corresponding to the target content to the message queue corresponding to the priority information; Taking out the push task of the target content from the message queue corresponding to the priority information; The allocating the candidate push device to the target machine node comprises: Based on the push task of the target content, the candidate push device is assigned to the target machine node.

10. The method according to claim 1, characterized in that The step of performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine the target push device from the assigned candidate push devices includes: For each candidate push device assigned to the target machine node, determine a push record logic block of the candidate push device, wherein the push record logic block includes at least one bit, and state identification information of the bit represents interaction information corresponding to the candidate push device and the content; Performing a hash operation on the target content to obtain a content hash value; Mapping the target content to the push record logic block based on the content hash value, and determining a mapping position of the target content in the push record logic block; The candidate push devices are filtered according to the status identification information of the bit corresponding to the mapping position to obtain the target push device.

11. The method according to claim 10, characterized in that The filtering the candidate push devices according to the state identification information of the bit corresponding to the mapping position to obtain the target push device includes: When the status identification information of each bit corresponding to the mapping position indicates that each bit has been set, filtering out the candidate push device; When the status identification information of each bit corresponding to the mapping position indicates that there is an unset bit among the bits, the candidate push device is retained, and the status identification information of the bit is updated to be set, so as to obtain the target push device corresponding to the target machine node.

12. The method according to claim 1, characterized in that The step of performing a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node to determine the target push device from the assigned candidate push devices includes: For each candidate push device assigned to the target machine node, obtaining historical push data of the candidate push device in at least one dimension through the target machine node; When the historical push data in the at least one dimension meets a preset condition, the candidate push device is determined as a target push device.

13. A content push device, characterized in that: include: A determination unit, used to determine the target content to be pushed and the push object group identifier corresponding to the target content; and determining at least one machine node for pushing the target content; A recall unit, configured to recall at least one push device set corresponding to the target content based on the push object group identifier and the device set corresponding to the preset object group identifier, wherein the push device set includes at least one candidate push device; a node determination unit, configured to determine, for each candidate push device, a target machine node for processing the candidate push device from among the machine nodes based on an association relationship between device information of the candidate push device and mapping points of node information of each machine node in a preset data space, and assign the candidate push device to the target machine node; An analyzing unit, configured to perform a business demand analysis on the candidate push devices assigned to the target machine node through the target machine node, so as to determine a target push device from the assigned candidate push devices; A pushing unit is used to push the target content to the target pushing device.

14. An electronic device, characterized in that: It comprises a memory and a processor; the memory stores an application program, and the processor is used to run the application program in the memory to perform the operations in the content push method according to any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the content pushing method according to any one of claims 1 to 12.

16. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps in the content push method described in any one of claims 1 to 12 are implemented.