Information flow loading system, information flow loading timing data processing method and device

Abstract

This application discloses an information stream loading system, an information stream loading timing data processing method and apparatus. The system includes a first edge node server, a client, and a cloud server. The client sends browsing confidence data and network quality index data to the first edge node server. The first edge node server obtains first load index data, and based on the browsing confidence data, network quality index data, and first load index data, obtains and sends loading timing data to the client, and obtains and sends second load index data to the cloud server. The client sends a loading request to the first edge node server based on the loading timing data. The first edge node server provides the information stream to be loaded to the client based on the loading request. The cloud server adjusts the information stream loading timing data of at least one client when a trigger condition is met, and adjusts the information stream loading method when the second load index data is unacceptable. This solution can reduce server performance consumption.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to an information stream loading system, an information stream loading timing data processing method and apparatus. Background Technology

[0002] With the continuous development of computer technology, accessing content through browsing news feeds has gradually become the mainstream method. However, loading delays and stuttering during this process can lead to low user satisfaction and negatively impact user retention. To improve the smoothness of news feed browsing and reduce data loading time, the industry primarily employs content preloading. However, current content preloading methods lack flexibility, resulting in excessive server performance consumption and wasted server resources. Summary of the Invention

[0003] This application provides an information stream loading system, an information stream loading timing data processing method and apparatus, which can reduce server performance consumption and avoid wasting server resources. The specific solution is as follows:

[0004] In a first aspect, embodiments of this application provide an information stream loading system, comprising: a first edge node server, a client, and a cloud server; the client is configured to send information stream browsing confidence data, representing the confidence level of a user browsing an information stream to be loaded in the client, and network quality index data, representing the network quality of the client, to the first edge node server, wherein the information stream to be loaded is an information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task; the first edge node server is configured to obtain first load index data representing the first load state of the first edge node server, and load the information stream browsing confidence data and the network quality index data according to the information stream browsing confidence data and the network quality index data. The client, using the first load metric data and the first load metric data, obtains information stream loading timing data indicating the timing of loading the information stream to be loaded in the client, and sends the information stream loading timing data to the client; it also obtains second load metric data indicating the second load state of the first edge node server after loading the information stream to be loaded for the client according to the information stream loading timing data, and sends the second load metric data to the cloud server; the client is further configured to send an information stream loading request message to the first edge node server according to the information stream loading timing data; the first edge node server is further configured to send an information stream loading request message to the client according to the information stream loading request message. The cloud server provides the information stream to be loaded. It determines whether the triggering condition for adjusting the information stream loading timing data of the client served by the first edge node server is met. If so, it adjusts the information stream loading timing data of at least one client served by the first edge node server so that the loading timing adjustment objective function value is an acceptable loading timing adjustment objective function value. The loading timing adjustment objective function includes a seamless loading parameter, an information stream usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception of the information stream being loaded in the client served by the first edge node server. The information stream usage parameter represents the user's usage of the information stream loaded in the client served by the first edge node server. The resource consumption parameter represents the resource consumption state of the first edge node server when loading the information stream in the client served by the first edge node server. The values ​​of the seamless loading parameter and the information stream usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. The cloud server also determines whether the second load indicator data is acceptable load indicator data. If the second load indicator data is not acceptable load indicator data, it adjusts the information stream loading method of the first edge node server.

[0005] Secondly, embodiments of this application provide a method for processing information stream loading timing data, comprising: acquiring information stream browsing confidence data representing a user browsing an information stream to be loaded in a client, wherein the information stream to be loaded is an information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task; acquiring network quality index data representing the network quality of the client; acquiring first load index data representing the load status of a first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client; and obtaining information stream loading timing data representing the timing of loading the information stream to be loaded in the client based on the information stream browsing confidence data, the network quality index data, and the first load index data.

[0006] Thirdly, embodiments of this application provide a method for processing information stream loading timing data, comprising: in response to detecting that a user stops swiping in the graphical user interface of a client, obtaining information stream browsing confidence data representing the user browsing an information stream to be loaded in the client, wherein the information stream to be loaded is an information stream that is not loaded in the client but needs to be loaded in the next information stream loading task; obtaining network quality index data representing the network quality of the client; sending the information stream browsing confidence data and the network quality index data to a first edge node server, and receiving information stream loading timing data returned by the first edge node server, which is obtained based on the information stream browsing confidence data, the network quality index data, and a first load index data representing the load status of the first edge node server, representing the timing of loading the information stream to be loaded in the client.

[0007] Fourthly, embodiments of this application provide a method for processing information stream loading timing data, comprising: determining whether a triggering condition for adjusting the information stream loading timing data of a client served by a first edge node server is met; if so, adjusting the information stream loading timing data of at least one client served by the first edge node server so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value; the loading timing adjustment objective function includes a seamless loading parameter, an information stream usage parameter, and a resource consumption parameter, wherein the seamless loading parameter represents the user's perception state of loading an information stream in a client served by the first edge node server, the information stream usage parameter represents the user's usage state of the information stream loaded in the client served by the first edge node server, and the resource consumption parameter represents the resource consumption state of the first edge node server when loading an information stream in a client served by the first edge node server, wherein the values ​​of the seamless loading parameter and the information stream usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function; and sending the adjusted information stream loading timing data of at least one client served by the first edge node server to the first edge node server.

[0008] Fifthly, embodiments of this application provide a first information stream loading timing data processing apparatus, comprising: a first acquisition unit, configured to acquire information stream browsing confidence data representing a user browsing an information stream to be loaded in a client, wherein the information stream to be loaded is an information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task; a second acquisition unit, configured to acquire network quality index data representing the network quality of the client; a third acquisition unit, configured to acquire first load index data representing the load status of a first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client; and a first obtaining unit, configured to obtain information stream loading timing data representing the timing of loading the information stream to be loaded in the client based on the information stream browsing confidence data, the network quality index data, and the first load index data.

[0009] In a sixth aspect, embodiments of this application provide a second information stream loading timing data processing apparatus, comprising: a second obtaining unit, configured to obtain information stream browsing confidence data representing the user browsing an information stream to be loaded in the client in response to detecting that a user stops swiping in the graphical user interface of a client, wherein the information stream to be loaded is an information stream that is not loaded in the client but needs to be loaded in the next information stream loading task; a third obtaining unit, configured to obtain network quality index data representing the network quality of the client; and a receiving unit, configured to send the information stream browsing confidence data and the network quality index data to a first edge node server, and receive information stream loading timing data returned by the first edge node server, which is obtained based on the information stream browsing confidence data, the network quality index data, and a first load index data representing the load status of the first edge node server, representing the timing of loading the information stream to be loaded in the client.

[0010] In a seventh aspect, embodiments of this application provide a third information stream loading timing data processing apparatus, comprising: a judgment unit, configured to determine whether a triggering condition for adjusting the information stream loading timing data of a client served by a first edge node server is met; and an adjustment unit, configured to adjust the information stream loading timing data of at least one client served by the first edge node server when the triggering condition for adjusting the information stream loading timing data of a client served by the first edge node server is met, so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value; the loading timing adjustment objective function includes a seamless loading parameter, an information stream usage parameter, and a resource consumption parameter, wherein the seamless loading parameter represents the user's perception of the information stream loading timing data of the client served by the first edge node server. The system is configured to: 1) perceive the loading status of information streams in clients served by the first edge node server; 2) use the information stream parameter to indicate the user's usage status of the information stream loaded in the clients served by the first edge node server; and 3) use the resource consumption parameter to indicate the resource consumption status of the first edge node server when loading information streams in the clients served by the first edge node server. The values ​​of the seamless loading parameter and the information stream usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. 4) send unit, configured to send adjusted information stream loading timing data for at least one client served by the first edge node server to the first edge node server.

[0011] Eighthly, this application also provides an electronic device, including: a processor; and a memory for storing a data processing program, wherein the electronic device, when powered on and running the program through the processor, performs the methods described in the second, third, or fourth aspects.

[0012] In a ninth aspect, embodiments of this application also provide a computer-readable storage medium storing a data processing program that is executed by a processor to perform the methods described in the second, third, or fourth aspects.

[0013] Compared with the prior art, this application has the following advantages:

[0014] In the information stream loading system provided in this application embodiment, the client is used to send information stream browsing confidence data, representing the user's confidence in browsing the information stream to be loaded in the client, and network quality index data, representing the network quality of the client, to the first edge node server. The information stream to be loaded is an information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task. The first edge node server is used to obtain first load index data, representing the first load state of the first edge node server, and obtain information stream loading timing data, representing the timing of loading the information stream to be loaded in the client, based on the information stream browsing confidence data, the network quality index data, and the first load index data, and send the information stream loading timing data to the client. The client is also used to send an information stream loading request message to the first edge node server based on the information stream loading timing data. The first edge node server is also used to provide the information stream to be loaded to the client based on the information stream loading request message. Because different clients may have different information stream browsing confidence data and network quality index data, and the first load index data of the first edge node server may differ at different times, and because the information stream browsing confidence data is used to measure the probability that a user will continue browsing the information stream to be loaded, and the network quality index data and the first load index data are used to measure the loading speed of the information stream to be loaded, by comprehensively considering the information stream browsing confidence data, the network quality index data, and the first load index data of the first load state, the corresponding information stream loading timing data can be determined flexibly and specifically for the clients served by the first edge node server. In this way, the client can flexibly load the information stream to be loaded based on the information stream loading timing data, avoiding excessive server performance consumption and resource waste. Attached Figure Description

[0015] Figure 1 This is a system architecture diagram of the information flow loading system provided in the embodiments of this application.

[0016] Figure 2 This is a signaling flowchart illustrating the communication and interaction between the first edge node server, the client, and the cloud server in the information flow loading system provided in this application embodiment.

[0017] Figure 3 This is a flowchart of the information flow loading timing data processing method provided in the second embodiment of this application.

[0018] Figure 4 This is a flowchart of the information flow loading timing data processing method provided in the third embodiment of this application.

[0019] Figure 5 This is a structural block diagram of an example of a data processing device for loading a first information stream, provided in an embodiment of this application.

[0020] Figure 6 This is a structural block diagram of an example of the second information stream loading timing data processing device provided in the embodiments of this application.

[0021] Figure 7 This is a structural block diagram of an example of an electronic device for data processing during information stream loading, provided in an embodiment of this application. Detailed Implementation

[0022] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0023] It should be noted that the terms "first," "second," "third," etc., in the claims, specification, and drawings of this application are used to distinguish similar objects and are not used to describe a specific order or sequence. Such data are interchangeable where appropriate so that the embodiments of this application described herein can be implemented in a sequence other than that shown or described herein. Furthermore, the terms "comprising," "having," and their variations are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or apparatuses.

[0024] It should be understood that in the embodiments of this application, "at least one" means one or more, and "more than one" means two or more. "And / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "Containing A, B, and / or C" means containing any one, two, or three of A, B, and C.

[0025] It should be understood that in the embodiments of this application, "B corresponding to A", "B corresponding to A", "A corresponds to B" or "B corresponds to A" means that B is associated with A, and B can be determined based on A. Determining B based on A does not mean that B is determined solely based on A; B can also be determined based on A and / or other information.

[0026] Before describing the implementation methods of this application in detail, the prior art will be further explained first.

[0027] Current content preloading methods use a fixed and uniform standard for all users. For example, for each user browsing their information stream, content preloading begins when there are only five items left to browse. However, on the one hand, due to individual user differences, some users may not continue browsing, leading to excessive waste of server resources if the server continuously preloads content. On the other hand, due to varying network conditions, using the same preloading standard (such as preloading only when there are only five items left) when the network speed is faster will result in the loaded content occupying a large amount of storage space on the server. Therefore, the current preloading method suffers from poor flexibility, excessive server performance consumption, and waste of server resources.

[0028] For the reasons stated above, the first embodiment of this application provides an information stream loading system to reduce server performance consumption and avoid wasting server resources. To more clearly demonstrate the information stream loading system provided in this embodiment, the application scenarios of the information stream loading system provided in this embodiment are first introduced.

[0029] The information flow loading system provided in this application can be applied to transaction and delivery services on instant e-commerce platforms, such as Taobao Flash Sale, Taoxianda, Ele.me food delivery, and retail. Specifically, it can be used to load product recommendation lists or store recommendation lists for consumers on instant e-commerce platforms. The information flow loading system provided in this application can also be applied to instant social platforms. Specifically, it can be used to load social information from instant social platforms for users. The information flow loading system provided in this application can also be applied to online information retrieval platforms. Specifically, it can be used to load information from online information retrieval platforms for users. It should be noted that the application scenarios given above are merely examples. Those skilled in the art can apply the information flow loading system provided in this application to other scenarios according to actual needs, and this application will not elaborate on each one.

[0030] To facilitate understanding of the information flow loading system provided in the embodiments of this application, the following is combined with... Figure 1This application describes the system architecture of the information stream loading system 100 provided in this embodiment. The information stream loading system 100 may include a cloud server 103, an edge node server cluster including a first edge node server 102, clients including at least one client 101, and a network 104. The client 101 is a client served by the first edge node server 102. The client 101 can be any device with computing hardware capable of supporting and executing the software application tools corresponding to the information stream loading system. Client 101 can be an electronic device with display and data processing capabilities, such as a mobile phone, tablet, desktop computer, or laptop. Users can browse the information stream on client 101 by swiping. In response to detecting a user stopping the swiping action on the information stream display interface of client 101, client 101 sends information stream browsing confidence data, representing the user's confidence in browsing the information stream to be loaded, and network quality index data, representing the network quality of the client, to first edge node server 102. It also receives information stream browsing loading timing data returned by first edge node server 102. When the timing for loading the information stream to be loaded, as indicated by the information stream browsing loading timing data, client 101 sends an information stream loading request message to first edge node server 102 to provide the information stream to be loaded to client 101 through first edge node server 102. First edge node server 102 refers to a distributed server node deployed near the network "edge" of client 101, typically located near a regional data center, carrier server room, CDN (Content Delivery Network) node, or local base station. The first edge node server 102 has a network connection with the client 101. Upon receiving information stream browsing confidence data and network quality indicator data sent by the client 101, it calculates information stream browsing loading timing data for the client 101 and sends this data back. Upon receiving an information stream loading request message from the client 101, it provides the information stream to be loaded to the client 101. The first edge node server 102 is physically closer to the client 101, enabling localized processing and response to request messages from the client 101, thereby reducing network latency and alleviating the burden on the cloud server 103. The cloud server 103 refers to a virtualized or physical server deployed in a centralized, large-scale data center. As a core component of the cloud computing infrastructure, it is responsible for providing powerful computing, storage, network resources, and global logical processing capabilities. The cloud server 103 and the edge node server closer to the client form a "center-edge" collaborative architecture.The cloud server 103 has network connections with multiple edge node servers (such as the first edge node server 102) in the edge node server cluster, used for global monitoring of the load metric data of the edge node servers (such as the first edge node server 102). The network 104 can be a wireless network or a wired network, such as a wireless local area network (WLAN), local area network (LAN), cellular network, 2G network, 3G network, 4G network, 5G network, etc. It should be noted that... Figure 1 The schematic diagram of the information flow loading system shown is merely an example. The information flow loading system 100 described in this application embodiment is intended to more clearly illustrate the technical solution of this application embodiment and does not constitute a limitation on the technical solution provided in this application embodiment.

[0031] The technical solution of this application will be described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0032] The first embodiment of this application provides an information stream loading system, such as... Figure 2 The diagram shown is a signaling flowchart of the communication interaction between the first edge node server, the client, and the cloud server in the information flow loading system provided in this application embodiment.

[0033] The client 101 is used to execute step S102: sending information stream browsing confidence data and network quality index data to the first edge node server 102. By executing step S102, the client 101 enables the first edge node server 102 to obtain the information stream browsing confidence data and network quality index data currently corresponding to the client 101 it serves, providing subjective user browsing metrics and objective network quality metrics for determining the information stream loading timing data for the client 101.

[0034] The client 101 is any one of the multiple clients served by the first edge node server 102.

[0035] The information stream browsing confidence data represents the confidence level of a user holding client 101 when browsing the information stream to be loaded on client 101. The information in the information stream to be loaded is recommended to the user based on the user's user behavior data and matches the user's user preferences. In an instant e-commerce platform, the information stream to be loaded may include a preset number (e.g., 15, 20, or 25) of product recommendation lists or store recommendation lists. Different users can correspond to different information streams to be loaded; for example, user A's information stream may include 20 parenting-related store recommendations, user B's may include 20 pet-related store recommendations, and user C's may include 20 sports and fitness-related store recommendations.

[0036] The confidence level of a user browsing the pending information stream in client 101 refers to the probability that the user will perform a swipe operation in client 101 to switch the content displayed in the information stream display interface of client 101 to the pending information stream. In other words, the information stream browsing confidence level data represents the probability that a user will perform a swipe operation in client 101 to switch the content displayed in the information stream display interface of client 101 to the pending information stream. In practical applications, the information stream browsing confidence level data may differ for different users.

[0037] In an optional implementation, the client 101 can obtain the information stream browsing confidence data by executing step S101. Step S101: In response to detecting the user's stop scrolling operation on the information stream display interface of the client 101, obtain information stream browsing confidence data and network quality index data.

[0038] Specifically, when the user scrolls through the information stream on the client 101's information stream display interface, they can stop scrolling when they reach a certain piece of information. At this time, the client 101 can obtain the information stream browsing confidence data based on the scrolling distance prediction data (which represents the user's predicted scrolling distance on the client 101's information stream display interface) and the scrolling speed prediction data (which represents the user's predicted scrolling speed on the client 101's information stream display interface). In specific implementation, the sliding trajectory prediction model preset in the client 101 can predict the sliding distance prediction data and the sliding speed prediction data based on sliding acceleration curve data (representing the rate of change of sliding speed when the user performs sliding operations to browse the historical information stream in the information stream display interface of the client 101), touch pressure data (representing the touch pressure when the user performs sliding operations to browse the historical information stream in the information stream display interface of the client 101), and historical information stream browsing status data (representing the historical browsing status when the user performs sliding operations to browse the historical information stream in the information stream display interface of the client 101). Then, the information stream browsing confidence data can be obtained based on the sliding distance prediction data and the sliding speed prediction data. The historical information stream browsing status data may include, but is not limited to, the browsing time data of each piece of content in the historical information stream, the user feedback data of each piece of content in the historical information stream, and the proportion data of each information type in the historical information stream.

[0039] In another implementation, the information flow browsing confidence data can be obtained based on the target historical information flow browsing status data, which represents the browsing status of the user when performing a swiping operation to browse the target historical information flow in the information flow display interface of the client 101 within the target time period. The target time period is the time period between the current time and the target time, and the target time is the time period before the current time and the time period between the current time and the current time is a preset time period (such as 5 days, 10 days or 15 days).

[0040] Thus, by obtaining the information stream browsing confidence data, the probability of a user continuing to browse the information stream to be loaded on the client 101 can be determined. In this way, the information stream loading timing data can be determined specifically for the client 101, avoiding the problem of excessive server performance consumption and resource waste caused by using the same information stream loading timing for all clients served by the first edge node server 102.

[0041] The network quality index data is used to represent the network quality of the client 101 at the current moment, and the network quality is used to describe the stability and efficiency of the network when transmitting data.

[0042] In an optional implementation, the client can obtain the network quality index data by performing the above step S101, that is, in response to detecting a user's stop scrolling operation in the information flow display interface of the client 101, the client obtains the network quality index data. Specifically, during the process of the user scrolling through the information flow in the information flow display interface of the client 101, the user can stop scrolling when browsing a certain piece of information. At this time, the client 101 can obtain the network quality index data based on at least one of the following: information flow loading duration data indicating the time consumed by the client 101 in the previous information flow loading task, network connection type data indicating the network type connected to the client 101, and geographical location type data indicating the geographical location type of the client 101.

[0043] In one optional specific example, the network deviation degree of the current network quality of client 101 relative to the preset network quality can be obtained based on the information stream loading time data representing the time consumed by client 101 in loading the information stream in the previous information stream loading task and preset information stream loading time data. The network quality index data is then determined based on the network deviation degree. Specifically, determining the network quality index data based on the network deviation degree can involve: determining the network deviation degree interval in which the network deviation degree falls, and determining the network quality index data corresponding to the network deviation degree interval as the network quality index data corresponding to the network deviation degree. The information stream loading time data is negatively correlated with network quality; that is, the longer the duration represented by the information stream loading time data, the worse the network quality of client 101. Since the network state of client 101 usually does not differ significantly between two adjacent information stream loading tasks, the current network quality of client 101 can be determined relatively accurately and efficiently using the information stream loading time data representing the time consumed by client 101 in loading the information stream in the previous information stream loading task, thereby obtaining the network quality index data.

[0044] In another optional specific example, the network quality index data can be obtained based on network connection type data, which indicates the network type to which client 101 is connected. The network connection type data may include data indicating that client 101 is connected to a 5G network, a 4G network, a 3G network, a 2G network, a 1G network, or has no network connection, etc. Specifically, different network connection type data can correspond to different network quality scores. For example, a 5G network quality score is 20 points, a 4G+ network quality score is 16 points, a 4G network quality score is 12 points, a 3G network quality score is 8 points, a 2G network quality score is 4 points, a 1G network quality score is 1 point, and a network quality score of 0 points indicates no network connection. Thus, the network quality score corresponding to the network connection type data can be determined as the network quality index data. Since different network connection types correspond to different data transmission speeds, generally speaking, the data transmission speed of 5G network is greater than that of 4G+ network, the data transmission speed of 4G+ network is greater than that of 4G network, the data transmission speed of 4G network is greater than that of 3G network, the data transmission speed of 3G network is greater than that of 2G network, the data transmission speed of 2G network is greater than that of 1G network, and the data transmission speed of 1G network is greater than that of no network connection. In this way, the current network quality of the client 101 can be accurately obtained based on the network connection type data, thereby obtaining the network quality index data.

[0045] In another optional specific example, the network quality index data can be obtained based on the geographic location type data representing the geographic location type of the client 101. The geographic location type can be a geographic location type divided according to the base station density within a preset spatial range of the client 101's geographic location and / or the location type of the client 101's geographic location. For example, the geographic location type may include, but is not limited to, a geographic location type representing the client 101 being located in a subway, a geographic location type representing the client 101 being located in a shopping mall, a geographic location type representing the client 101 being located in an elevator, and a geographic location type representing the client 101 being located in an open outdoor area, etc. Since the network quality varies in different geographic location types—for example, the network quality is poor in subways, elevators, and underground parking garages, while the network quality is good in open outdoor areas—the network quality corresponding to the client 101 can be accurately obtained based on the geographic location type data representing the client 101's geographic location type, thereby obtaining the network quality index data.

[0046] In another alternative implementation, the client 101 may also predict the network quality index data in the previous information stream loading task.

[0047] In related technologies, all clients served by the first edge node server 102 use the same information stream loading timing. This can lead to issues such as clients with good network quality completing loading quickly, causing the loaded information stream to occupy memory for a long time, while clients with poor network quality cannot complete loading in time, resulting in users seeing a loading wait process. This solution obtains the network quality indicator data. Since the network quality of client 101 directly affects the data loading speed of client 101, the information stream loading timing data can be specifically determined for client 101. This avoids excessive server performance consumption and wasted server resources.

[0048] After obtaining the information stream browsing confidence data and the network quality index data, the client 101 sends the information stream browsing confidence data and the network quality index data to the first edge node server 102.

[0049] After receiving the information stream browsing confidence data and the network quality index data, the first edge node server 102 executes step S103: acquiring first load index data representing the first load state of the first edge node server 102. The first load state refers to the current resource utilization of the first edge node server 102, used to measure the current performance of the first edge node server 102. The first load state can be any one of idle state, normal operation state, overload state, and downtime state. The first load state of the first edge node server 102 may differ at different times. Taking the information stream loading system applied to an instant e-commerce platform as an example, typically, 11:00~13:30 and 17:00~20:00 are peak periods, and the first load state may be overloaded; 23:00~6:00 the next day is an off-peak period, and the first load state may be idle. Under different load conditions, the data processing speed of the first edge node server 102 varies. The data processing speed of the first edge node server 102 in idle state is greater than or equal to its data processing speed in normal operation state. The data processing speed of the first edge node server 102 in normal operation state is greater than or equal to its data processing speed in overload state. The data processing speed of the first edge node server 102 in overload state is greater than or equal to its data processing speed in a downtime state. The first load metric data may include, but is not limited to, at least one of the following: CPU utilization, memory utilization, network bandwidth utilization, and the number of processes processed.

[0050] Since the information stream to be loaded by the client 101 is provided by the first edge node server 102, the first load index data of the first edge node server 102 can affect the speed at which the information stream is loaded for the client 101. Thus, by obtaining the first load index data of the first edge node server 102 serving the client 101, an objective dimension of server load measurement can be provided for determining the information stream loading timing data for the client 101. In this way, the information stream browsing confidence data and the network quality index data can be combined to accurately obtain the information stream loading timing data.

[0051] After obtaining the information stream browsing confidence data, the network quality index data, and the first load index data, the first edge node server 102 can execute step S104: based on the information stream browsing confidence data, the network quality index data, and the first load index data, obtain information stream loading timing data to indicate the timing of loading the information stream to be loaded in the client.

[0052] The information stream loading timing data is either first information stream loading timing data in the time dimension or second information stream loading timing data in the spatial dimension. Specifically, the first information stream loading timing data indicates the time difference between the moment when the client 101 loads the information stream to be loaded and the moment when the user browses the information stream on the client 101; that is, the moment when the client 101 loads the information stream to be loaded is 10 seconds earlier than the moment when the user browses the information stream on the client 101. The second information stream loading timing data indicates the spatial distance between the interface display data when the client 101 loads the information stream to be loaded and the interface display data when the user browses the information stream on the client 101; the interface display data is at least one of the displayed point-of-interest (POI) card data and the displayed pixel data. Specifically, the first information stream loading timing data is used to indicate the number of store cards between the store cards displayed at the target card display position on the interface when the client 101 loads the information stream to be loaded and the store cards displayed at the target card display position on the interface when the user browses the information stream to be loaded on the client 101; or the first information stream loading timing data is used to indicate the number of pixels between the pixels displayed at the target pixel position on the interface when the client 101 loads the information stream to be loaded and the pixels displayed at the target pixel position on the interface when the user browses the information stream to be loaded on the client 101.

[0053] In one optional implementation, for all clients served by the first edge node server 102, the first information stream loading timing data can be used as the information stream loading timing data, or the second information stream loading timing data can be used as the information stream loading timing data.

[0054] In another optional implementation, for some clients served by the first edge node server 102, the first information stream loading timing data can be used as the information stream loading timing data; for another portion of the clients served by the first edge node server 102, the second information stream loading timing data can be used as the information stream loading timing data. In this implementation, the first edge node server 102 is further configured to acquire scrolling speed fluctuation data representing the historical scrolling speed fluctuation state of the user in the information stream display interface of the client. If the scrolling speed fluctuation data is less than a preset scrolling speed fluctuation threshold, then the first information stream loading timing data is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data, and used as the information stream loading timing data; otherwise, the second information stream loading timing data is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data, and used as the information stream loading timing data.

[0055] Specifically, if the historical scrolling speed fluctuation data of the user in the information feed display interface of the client is less than a preset scrolling speed fluctuation threshold, it indicates that the user's scrolling speed does not change much when browsing the information feed. In this case, the accuracy of predicting the first information feed loading timing data in the time dimension is high, and the first information feed loading timing data in the time dimension can be obtained as the information feed loading timing data. If the historical scrolling speed fluctuation data of the user in the information feed display interface of the client is greater than or equal to the preset scrolling speed fluctuation threshold, it indicates that the user's scrolling speed changes significantly when browsing the information feed. In this case, since the scrolling speed is uncontrollable, the accuracy of predicting the first information feed loading timing data in the time dimension is low. Therefore, the first information feed loading timing data in the spatial dimension can be obtained as the information feed loading timing data. This improves the accuracy of the information feed loading timing data prediction.

[0056] In one optional implementation, the information flow browsing confidence data, the network quality index data, and the first load index data can be used as the values ​​of the independent variables of a preset information flow loading timing data acquisition function, and input into the information flow loading timing data acquisition function to obtain the information flow loading timing data. The information flow loading timing data acquisition function can be a linear regression function. A linear regression function is used to predict a continuous numerical output, and its goal is to find a fitted straight line or a hyperplane in a multidimensional space. This fitted straight line or hyperplane is used to describe the linear relationship between the input variables (in this embodiment, the information flow browsing confidence data, the network quality index data, and the first load index data) and the output variables (in this embodiment, the information flow loading timing data).

[0057] In another optional implementation, the information flow loading timing data can be predicted using a trained information flow loading timing data prediction model. Specifically, the information flow browsing confidence data, the network quality index data, and the first load index data can be input into the trained information flow loading timing data prediction model, and the model can output the information flow loading timing data.

[0058] Since the information stream browsing confidence data and network quality index data may differ for different clients, and the first load index data of the first edge node server may differ at different times, and since the information stream browsing confidence data is used to measure the probability that a user will continue to browse the information stream to be loaded, and the network quality index data and the first load index data are used to measure the loading speed of the information stream to be loaded, by comprehensively considering the information stream browsing confidence data, the network quality index data, and the first load index data of the first load state, the corresponding information stream loading timing data can be determined for the clients served by the first edge node server in a targeted and flexible manner.

[0059] After obtaining the information stream loading timing data, the first edge node server 102 can execute step S105: send the information stream loading timing data to the client 101. The client 101 executes step S106: receive the information stream loading timing data, and after receiving the information stream loading timing data, executes step S107: send an information stream loading request message to the first edge node server 102 according to the information stream loading timing data. Step S107 is used to send an information stream loading request message to the first edge node server 102, so as to provide the information stream to be loaded to the client 101 through the first edge node server 102.

[0060] The first edge node server 102 provides the information stream to be loaded to the client 101 by executing steps S108 and S109. Step S108: Receive the information stream loading request message. Step S109: Provide the information stream to be loaded according to the information stream loading request message. Specifically, step S109 involves: the first edge node server 102 providing the client 101 with the information stream to be loaded stored on the first edge node server 102; or the first edge node server 102 sending the information stream loading request to the cloud server 103 and providing the information stream to be loaded returned by the cloud server 103 to the client 101.

[0061] In practical applications, when loading the information stream to be loaded onto the first edge node server 102 according to the information stream loading data, the server may become overloaded. In this case, the server's CPU can be increased, which will greatly increase the server cost. To solve this problem, the information stream loading system provided in this application embodiment can further monitor and adjust the information stream loading timing data.

[0062] Specifically, the first edge node server 102 can execute step S110: acquiring second load indicator data; the second load indicator data is data representing the second load state of the first edge node server 102 after loading the information stream to be loaded for the client 101 according to the information stream loading timing data. The second load state can be any one of idle state, normal operation state, overload state, and downtime state. In a specific implementation, the first edge node server 102 can acquire the second load indicator data according to a preset statistical period or in real time. The preset statistical period can be, for example, every 5 minutes, every 10 minutes, or every 15 minutes, etc., and can be specifically set based on actual needs. Then, the first edge node server 102 executes step S111: sending the second load indicator data to the cloud server 103.

[0063] After receiving the second load indicator data, the cloud server 103 can execute steps S112 and S113. Step S112: Determine whether the second load indicator data is acceptable load indicator data. Acceptable load indicator data is preset data, such as data indicating that the second load state of the first edge node server 102 is overloaded, or data indicating that the first edge node server 102 is operating at full load. Step S113: If the second load indicator data is not acceptable load indicator data, adjust the information flow loading method of the first edge node server.

[0064] In one optional implementation, step S113 includes: identifying a first client among the clients served by the first edge node server 102 that requires adjustment of the information stream loading timing data; adjusting the information stream loading timing data of the first client so that the third load indicator data representing the third load state of the first edge node server after loading the information stream to be loaded for the first client according to the adjusted information stream loading timing data is acceptable load indicator data. The first client is at least a portion of the clients served by the first edge node server 102.

[0065] This implementation method is used to determine the first client that needs to readjust the information flow loading timing data from the clients served by the first edge node server 102, and to change the unacceptable second load index data of the first edge node server to the acceptable third load index data by adjusting the information flow loading timing data of the first client (such as shortening the pre-loading time, reducing the number of pre-loaded cards, or reducing the number of pre-loaded pixels).

[0066] In a specific implementation, the first client can be determined in the following way: obtain historical behavior data of the clients served by the edge node server in performing historical information stream loading tasks; based on the historical behavior data, obtain information stream loading waiting state data representing the user's waiting state when the user perceives the information stream loading process in the clients served by the edge node server; based on the information stream loading waiting state data, determine the first client from the clients served by the first edge node server that needs to adjust the information stream loading timing data.

[0067] The historical behavior data refers to the data corresponding to the user's interactive behaviors on the client during the execution of the historical information stream loading task. The historical behavior data includes information stream loading wait state data, representing the user's waiting state when the information stream loading process is perceived in the clients served by the edge node server. This user waiting state may include, but is not limited to, at least one of the following: a first user waiting state of continuing to wait for information stream loading, and a second user waiting state of exiting the information stream display interface and no longer waiting for information stream loading. Subsequently, clients among the clients served by the first edge node server whose frequency of being in the first user waiting state when the information stream loading process is perceived in the clients served by the edge node server is greater than a preset frequency are identified as the first clients whose information stream loading timing data needs to be adjusted.

[0068] By analyzing the user's waiting status, the system can filter out the first client held by users who are more willing to wait for the information stream to load from the clients served by the first edge node server. In this way, by adjusting the information stream loading timing data of these users (such as shortening the pre-loading time, reducing the number of pre-loaded cards, or reducing the number of pre-loaded pixels), even if the information stream loading process is detected, users will not directly exit the information stream display interface, which is beneficial to user retention.

[0069] After identifying the first client, the cloud server 103 can obtain the target adjustment range data and adjust the information flow loading timing data of the first client according to the target adjustment range data, so that the third load index data representing the third load state of the first edge node server after loading the information flow to be loaded for the first client according to the adjusted information flow loading timing data is acceptable load index data.

[0070] Specifically, when the information stream loading timing data is the first information stream loading timing data, the information stream loading timing data of the first client is adjusted according to the target duration adjustment range data to obtain the adjusted information stream loading timing data. When the information stream loading timing data is the second information stream loading timing data, the information stream loading timing data of the first client is adjusted according to the target space adjustment range data to obtain the adjusted information stream loading timing data. After obtaining the adjusted information stream loading timing data of the first client, the cloud server 103 continues to perform the following steps: sending the adjusted information stream loading timing data to the first edge node server 102. The first edge node server 102 performs the following steps: sending the adjusted information stream loading timing data to the first client 101. Thus, when the first client 101 receives the adjusted information stream loading timing data, it can send an information stream loading request message to the first edge node server 102 to provide the information stream to be loaded to the first client through the first edge node server 102.

[0071] This configuration method, by adjusting the information flow loading timing data of the first client among the clients served by the first edge node server, enables the first edge node server to delay loading the information flow to be loaded for the first client. This avoids preloaded data occupying storage space for a long time, improves the performance of the first edge node server, and avoids resource waste of the first edge node server.

[0072] In another optional implementation, step S113 includes: determining a second client from among the clients served by the first edge node server that needs to be reassigned to an edge node server; determining a second edge node server from among the edge node servers associated with the cloud server that can serve the second client; and changing the edge node server serving the second client from the first edge node server to the second edge node server, so that after changing the edge node server serving the second client, the fourth load metric data representing the fourth load state of the first edge node server is acceptable load metric data. The second client is at least a portion of the clients served by the first edge node server 102. The second edge node server capable of serving the second client is the edge node server whose current fifth load metric data for the fifth load state is acceptable load metric data.

[0073] This implementation method is used to determine a second client that needs to be reassigned to an edge node server from the clients served by the first edge node server 102, and to change the edge node server serving the second client from the first edge node server to the second edge node server, so that the first edge node server changes from the unacceptable second load index data to the acceptable fourth load index data.

[0074] This configuration, by changing the edge node server serving the second client among the clients served by the first edge node server from the first edge node server to the second edge node server, enables the edge node servers to maintain load balance, avoid some edge node servers being overloaded and others being idle, improve the performance of the edge node servers, and avoid the waste of edge node servers resources.

[0075] In another optional implementation, step S113 includes: changing the compression method of the view information in the information stream to be loaded from a first compression method to a second compression method, and sending a view information compression instruction to the first edge node server, indicating that the view information in the information stream to be loaded should be compressed according to the second compression method; the compression loss of the second compression method is greater than the compression loss of the first compression method. The view information includes, but is not limited to, at least one of image information, video information, live streaming information, etc.

[0076] This implementation adjusts the compression method of the view information in the information stream to be loaded by the first edge node server 102. The compression method may include at least two of the following: lossless compression, low-loss compression, standard compression, and extreme compression. Lossless compression refers to compression where there is no information loss between the compressed file and the original file. Low-loss compression refers to compression where the compression loss between the compressed file and the original file is less than a first loss value. Standard compression refers to compression where the compression loss between the compressed file and the original file is greater than or equal to the first loss value and less than a second loss value. Extreme compression refers to compression where the compression loss between the compressed file and the original file is greater than or equal to the second loss value. If the first compression method is lossless, the second compression method can be any one of low-loss compression, standard compression, or extreme compression. If the first compression method is low-loss compression, the second compression method can be any one of standard compression or extreme compression. If the first compression method is standard compression, the second compression method can be extreme compression.

[0077] Compression methods with smaller compression loss require more complex pattern matching, dictionary construction, and entropy encoding, resulting in greater computational resource consumption. In this solution, when the second load index data of the first edge node server is unacceptable, the first compression method of the first edge node server for the view information in the information stream to be loaded is changed to a compression method with larger compression loss, which can reduce the computational resource consumption of the first edge node server during the information stream loading process.

[0078] When the first edge node server 102 receives the view information compression instruction, it can compress the view information in the information stream to be loaded according to the second compression method, obtain the compressed information stream to be loaded, and provide the compressed information stream to be loaded to the client 101.

[0079] In an optional implementation, the cloud server 103 is further configured to determine whether the triggering condition for adjusting the information flow loading timing data of the clients served by the first edge node server is met. If so, the information flow loading timing data of at least one client served by the first edge node server (e.g., clients numbered 3, 6, and 9 out of 100 clients) is adjusted so that the loading timing adjustment objective function value is an acceptable loading timing adjustment objective function value. The loading timing adjustment objective function includes a seamless loading parameter, an information flow usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception of the information flow being loaded in the clients served by the first edge node server. The information flow usage parameter represents the user's usage of the information flow loaded in the clients served by the first edge node server. The resource consumption parameter represents the resource consumption state of the first edge node server when the information flow is loaded in the clients served by the first edge node server. The values ​​of the seamless loading parameter and the information flow usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function.

[0080] The triggering condition for adjusting the information stream loading timing data of the clients served by the first edge node server can be determined by at least one of the following: time conditions, whether the load index data of the first edge node server is acceptable, and whether the value of the loading timing adjustment objective function is acceptable. In one example, the triggering condition for adjusting the information stream loading timing data of the clients served by the first edge node server can be a time condition, such as triggering the adjustment of the information stream loading timing data of the clients served by the first edge node server every five minutes. In another example, the triggering condition for adjusting the information stream loading timing data of the clients served by the first edge node server can be that the load index data of the first edge node server is unacceptable. In yet another example, the triggering condition for adjusting the information stream loading timing data of the clients served by the first edge node server can be that the value of the loading timing adjustment objective function is an unacceptable loading timing adjustment objective function value.

[0081] The seamless loading parameter (also known as the seamless loading rate) represents the user's perception of the information stream being loaded in the client served by the first edge node server. This parameter measures whether the information stream to be loaded has been pre-loaded when the user browses it. The information stream usage parameter (also known as the information stream usage rate) represents the user's usage status of the loaded information stream in the client served by the first edge node server. In an instant e-commerce platform, the user's usage status of the loaded information stream is used to characterize the number of products or stores consumed by the user. The resource consumption parameter (also known as the resource consumption rate) represents the resource consumption status of the first edge node server when loading the information stream in the client served by the first edge node server. Specifically, it refers to the resource consumption status of the first edge node server when loading the information stream in the client served by the first edge node server during a certain time period (e.g., 12:00~12:05, or 12:05~12:10).

[0082] The following formula (1) is the functional expression of the objective function for adjusting the loading timing:

[0083] Formula (1)

[0084] in, Adjust the value of the objective function for the loading timing. For the aforementioned seamless loading parameters, Parameters are used for the information flow. The resource consumption parameter is defined as follows. In this embodiment, when the triggering condition for adjusting the information flow loading timing data of the client served by the first edge node server is met, the information flow loading timing data of at least one client served by the first edge node server is adjusted. The value of the loading timing adjustment objective function is recalculated using the adjusted values ​​of the seamless loading parameter, the information flow usage parameter, and the resource consumption parameter. The value of the loading timing adjustment objective function is used as a constraint until the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value.

[0085] Since the seamless loading parameter is usually proportional to the resource consumption parameter, the larger the value of the seamless loading parameter, the more resources are consumed. Conversely, a larger value of the resource consumption parameter means that if the goal is simply to ensure that the information stream to be loaded is completed in advance when the user browses it, the performance of the first edge node server will be sacrificed. In this solution, a loading timing adjustment objective function is constructed using the seamless loading parameter, the information stream usage parameter, and the resource consumption parameter. The loading timing data of the information stream served by at least one client of the first edge node server is adjusted based on the constraint that the value of the loading timing adjustment objective function is an acceptable value. In this way, when the value of the loading timing adjustment objective function is an acceptable value, a relatively balanced state can be achieved, with a high user seamless loading rate, a high information stream usage rate, and a low resource consumption rate.

[0086] As described above, the information stream loading timing data can be obtained through the information stream loading timing data acquisition function. Ideally, when loading the information stream to be loaded for the client served by the first edge node server 102 according to the information stream loading timing data, the user holding the client served by the first edge node server 102 will not be aware of the information stream loading process. However, since the information stream loading timing data is predicted information stream loading timing data, inaccurate predictions may occur. Therefore, the client can also perform the following steps: obtain imperceptible loading data representing the user's perception state of loading the information stream in the client served by the first edge node server, sliding distance residual distribution data representing the actual sliding distance of the user in the information stream display interface of the client served by the first edge node server and sliding distance predicted data representing the predicted sliding distance of the user in the information stream display interface of the client served by the first edge node server, and information stream exposure data representing the information stream loaded in the client served by the first edge node server, and send the imperceptible loading data, the sliding distance residual distribution data, and the information stream exposure data to the first edge node server. The first edge node server is further configured to send the seamless loading data, the sliding distance residual distribution data, and the information stream exposure data to the cloud server; the cloud server is configured to determine whether the seamless loading data, the sliding distance residual distribution data, and the information stream exposure data are all acceptable data. If not, the cloud server adjusts the information stream loading timing data acquisition function until the seamless loading data, sliding distance residual distribution data, and information stream exposure data obtained based on the adjusted information stream loading timing data acquisition function are all acceptable data, and then sends the adjusted information stream loading timing data acquisition function to the first edge node server.

[0087] The seamless loading data refers to the values ​​of the aforementioned seamless loading parameters. The actual sliding distance data and the predicted sliding distance data are from the same round, and the sliding distance residual distribution data is used to represent the degree of difference between the actual sliding distance data and the predicted sliding distance data. The information stream exposure data is used to represent the user's browsing status of the information stream loaded in the client served by the first edge node server.

[0088] The cloud server determines whether the seamless loading data, the sliding distance residual distribution data, and the information stream exposure data are all acceptable data. If any one of them is unacceptable, for example, the seamless loading data is unacceptable, the sliding distance residual distribution data is unacceptable, or the information stream exposure data is unacceptable, then the information stream loading timing data acquisition function is adjusted until the seamless loading data, sliding distance residual distribution data, and information stream exposure data obtained based on the adjusted information stream loading timing data acquisition function are all acceptable data. Then, the adjusted information stream loading timing data acquisition function is sent to the first edge node server.

[0089] In this way, the function for obtaining the information stream loading timing data of the first edge node server can be iteratively updated, thereby improving the accuracy of obtaining the information stream loading timing data through the function. Therefore, in the information stream loading system provided in this embodiment, the client can flexibly load the information stream to be loaded based on the information stream loading timing data, avoiding excessive server performance consumption and resource waste.

[0090] The second embodiment of this application also provides a method for processing information stream loading timing data, applied to the first edge node server in the information stream loading system. For example... Figure 3The diagram shows a flowchart of the information stream loading timing data processing method provided in the second embodiment of this application, including steps S201 to S204. Step S201: Obtain information stream browsing confidence data representing a user's browsing of an information stream to be loaded in the client, wherein the information stream to be loaded is an information stream that is not loaded in the client but needs to be loaded in the next information stream loading task. Step S202: Obtain network quality index data representing the network quality of the client. Step S203: Obtain first load index data representing the load status of a first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client. Step S204: Based on the information stream browsing confidence data, the network quality index data, and the first load index data, obtain information stream loading timing data representing the timing of loading the information stream to be loaded in the client.

[0091] The third embodiment of this application also provides a method for processing information stream loading timing data, applied to the client in the information stream loading system. For example... Figure 4 The diagram shows a flowchart of an information stream loading timing data processing method provided in the third embodiment of this application, including steps S301 to S303. Step S301: In response to detecting that a user stops swiping in the client's graphical user interface, obtain information stream browsing confidence data representing the user's browsing of an information stream to be loaded in the client. The information stream to be loaded is an information stream that is not loaded in the client but needs to be loaded in the next information stream loading task. Step S302: Obtain network quality index data representing the network quality of the client. Step S303: Send the information stream browsing confidence data and the network quality index data to a first edge node server, and receive information stream loading timing data returned by the first edge node server, obtained based on the information stream browsing confidence data, the network quality index data, and first load index data representing the load status of the first edge node server, representing the timing of loading the information stream to be loaded in the client.

[0092] The fourth embodiment of this application also provides a method for processing information flow loading timing data, applied to the cloud server in the information flow loading system, comprising the following steps: determining whether the triggering condition for adjusting the information flow loading timing data of the client served by the first edge node server is met; if so, adjusting the information flow loading timing data of at least one client served by the first edge node server so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value; the loading timing adjustment objective function includes a seamless loading parameter, an information flow usage parameter, and a resource consumption parameter, wherein the seamless loading parameter represents the user's perception state of loading information flow in the client served by the first edge node server, the information flow usage parameter represents the user's usage state of the information flow loaded in the client served by the first edge node server, and the resource consumption parameter represents the resource consumption state of the first edge node server when loading information flow in the client served by the first edge node server, wherein the values ​​of the seamless loading parameter and the information flow usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function; and sending the adjusted information flow loading timing data of at least one client served by the first edge node server to the first edge node server.

[0093] For a detailed description of the information flow loading timing data processing method provided in the second, third and fourth embodiments of this application, please refer to the relevant description of the first embodiment of this application, which will not be repeated here.

[0094] Corresponding to the information stream loading timing data processing method provided in the second embodiment of this application, the fifth embodiment of this application also provides a first information stream loading timing data processing device, such as... Figure 5 As shown, the first information stream loading timing data processing device 500 includes: a first acquisition unit 501, used to acquire information stream browsing confidence data representing a user browsing an information stream to be loaded in a client, wherein the information stream to be loaded is an information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task; a second acquisition unit 502, used to acquire network quality index data representing the network quality of the client; a third acquisition unit 503, used to acquire first load index data representing the load status of a first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client; and a first obtaining unit 504, used to obtain information stream loading timing data representing the timing of loading the information stream to be loaded in the client based on the information stream browsing confidence data, the network quality index data, and the first load index data.

[0095] Corresponding to the information stream loading timing data processing method provided in the third embodiment of this application, the sixth embodiment of this application also provides a second information stream loading timing data processing device, such as... Figure 6 As shown, the second information stream loading timing data processing device 600 includes: a second obtaining unit 601, configured to obtain information stream browsing confidence data representing the user browsing an information stream to be loaded in the client in response to detecting that a user stops swiping in the graphical user interface of the client, wherein the information stream to be loaded is an information stream that is not loaded in the client but needs to be loaded in the next information stream loading task; a third obtaining unit 602, configured to obtain network quality index data representing the network quality of the client; and a receiving unit 603, configured to send the information stream browsing confidence data and the network quality index data to a first edge node server, and receive information stream loading timing data returned by the first edge node server, which is obtained based on the information stream browsing confidence data, the network quality index data, and a first load index data representing the load status of the first edge node server, representing the timing of loading the information stream to be loaded in the client.

[0096] Corresponding to the information stream loading timing data processing method provided in the fourth embodiment of this application, the seventh embodiment of this application also provides a third information stream loading timing data processing device. The third information stream loading timing data processing device includes: a judgment unit, used to determine whether the triggering condition for adjusting the information stream loading timing data of the client served by the first edge node server is met; and an adjustment unit, used to adjust the information stream loading timing data of at least one client served by the first edge node server when the triggering condition for adjusting the information stream loading timing data of the client served by the first edge node server is met, so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value; the loading timing adjustment objective function includes a seamless loading parameter, an information stream usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception of the information stream being loaded in the client served by the first edge node server. The information stream usage parameter represents the user's usage of the information stream loaded in the client served by the first edge node server. The resource consumption parameter represents the resource consumption status of the first edge node server when loading the information stream in the client served by the first edge node server. The values ​​of the seamless loading parameter and the information stream usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. The sending unit is used to send the adjusted information stream loading timing data of at least one client served by the first edge node server to the first edge node server.

[0097] Corresponding to the information stream loading timing data processing method provided in the second, third, or fourth embodiments of this application, the eighth embodiment of this application also provides an electronic device for the information stream loading timing data processing method. For example... Figure 7 As shown, the electronic device provided in this embodiment includes: a processor 701, a memory 703, a communication bus 704, and a communication interface 702; the memory 703 is used to store computer instructions for data processing, and when the computer instructions are read and executed by the processor 701, the information flow loading timing data processing method provided in the second, third, or fourth embodiment of this application is executed.

[0098] Corresponding to the information flow loading timing data processing method provided in the second, third, or fourth embodiment of this application, the ninth embodiment of this application provides a computer-readable storage medium storing a program for the information flow loading timing data processing method. This program is executed by a processor to perform the information flow loading timing data processing method provided in the second, third, or fourth embodiment of this application.

[0099] It should be noted that for detailed descriptions of the apparatus, electronic device, and computer-readable storage medium provided in the fifth, sixth, seventh, eighth, and ninth embodiments of this application, please refer to the relevant descriptions of the second, third, and fourth embodiments of this application, which will not be repeated here.

[0100] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory. Memory may include non-persistent storage in computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media. 1. Computer-readable media includes both permanent and non-persistent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information that can be accessed by a computing device. As defined in this application, computer-readable media does not include non-transitory computer-readable media, such as modulated data signals and carrier waves. 2. Those skilled in the art will understand that embodiments of this application can provide methods, systems, or computer program products. Therefore, embodiments of this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. 3. Embodiments of this application may involve the use of user data. In practical applications, user-specific personal data can be used within the scope permitted by applicable laws and regulations, provided it complies with the applicable laws and regulations of the country (e.g., with the user's explicit consent, and with actual notification to the user, etc.). Furthermore, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points shall be provided for users to choose to authorize or refuse.

[0101] Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims of this application.

Claims

1. An information stream loading system, characterized in that, include: First edge node server, client, and cloud server; The client is configured to send information stream browsing confidence data, which represents the confidence level of the user browsing the information stream to be loaded in the client, and network quality index data, which represents the network quality of the client, to the first edge node server. The information stream to be loaded is the information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task. The first edge node server is configured to acquire first load index data representing the first load state of the first edge node server, and based on the information flow browsing confidence data, the network quality index data, and the first load index data, obtain information flow loading timing data representing the timing of loading the information flow to be loaded in the client, and send the information flow loading timing data to the client. Obtain second load index data representing the second load state of the first edge node server after loading the information stream to be loaded for the client according to the information stream loading timing data, and send the second load index data to the cloud server; The client is also used to send an information stream loading request message to the first edge node server based on the information stream loading timing data; The first edge node server is also configured to provide the information stream to be loaded to the client according to the information stream loading request message; The cloud server is used to determine whether the triggering condition for adjusting the information flow loading timing data of the client served by the first edge node server is met. If so, the information flow loading timing data of at least one client served by the first edge node server is adjusted so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value. The loading timing adjustment objective function includes a seamless loading parameter, an information flow usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception state of loading information flow in the client served by the first edge node server. The information flow usage parameter represents the user's usage state of the information flow loaded in the client served by the first edge node server. The resource consumption parameter represents the resource consumption state of the first edge node server when loading information flow in the client served by the first edge node server. The values ​​of the seamless loading parameter and the information flow usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. The cloud server is also used to determine whether the second load indicator data is acceptable load indicator data. If the second load indicator data is not acceptable load indicator data, the information flow loading method of the first edge node server is adjusted.

2. The system according to claim 1, characterized in that, The information stream loading timing data is either a first information stream loading timing data in the time dimension or a second information stream loading timing data in the spatial dimension; wherein, the first information stream loading timing data is used to indicate the time difference between the moment when the client loads the information stream to be loaded and the moment when the user browses the information stream to be loaded on the client, and the second information stream loading timing data is used to indicate the spatial distance between the interface display data when the client loads the information stream to be loaded and the interface display data when the user browses the information stream to be loaded on the client; the interface display data is at least one of the displayed point of interest card data and the displayed pixel data.

3. The system according to claim 1, characterized in that, The adjustment of the information flow loading method of the first edge node server includes: determining the first client whose information flow loading timing data needs to be adjusted from the clients served by the first edge node server, and adjusting the information flow loading timing data of the first client so that the third load index data representing the third load state of the first edge node server after loading the information flow to be loaded for the first client according to the adjusted information flow loading timing data is acceptable load index data, and sending the adjusted information flow loading timing data to the first edge node server. The first edge node server is also used to send the adjusted information stream loading timing data to the first client.

4. The system according to claim 1, characterized in that, The adjustment of the information flow loading method of the first edge node server includes: determining a second client that needs to be reassigned to an edge node server from the clients served by the first edge node server; determining a second edge node server that can serve the second client from the edge node servers associated with the cloud server; changing the edge node server serving the second client from the first edge node server to the second edge node server, so that the fourth load index data representing the fourth load state of the first edge node server after changing the edge node server serving the second client is acceptable load index data.

5. The system according to claim 1, characterized in that, The client is also configured to respond to detecting a user's stop swiping operation in the client's information flow display interface, and obtain the information flow browsing confidence data based on swipe distance prediction data representing the user's predicted swipe distance in the client's information flow display interface and swipe speed prediction data representing the user's predicted swipe speed in the client's information flow display interface. Furthermore, the client is configured to obtain the network quality index data based on at least one of the following: information flow loading duration data representing the duration consumed by the client in the previous information flow loading task, network connection type data representing the network type to which the client is connected, and geographic location type data representing the geographic location type of the client.

6. The system according to claim 1, characterized in that, The step of obtaining information stream loading timing data, which indicates the timing of loading the information stream to be loaded in the client, based on the information stream browsing confidence data, the network quality index data, and the first load index data, includes: inputting the information stream browsing confidence data, the network quality index data, and the first load index data as values ​​of independent variables of a preset information stream loading timing data acquisition function into the information stream loading timing data acquisition function to obtain the information stream loading timing data; The client is further configured to acquire: imperceptible loading data representing the user's perception state of loading information stream in the client served by the first edge node server; actual sliding distance data representing the user's actual sliding distance in the information stream display interface of the client served by the first edge node server and sliding distance prediction data representing the user's predicted sliding distance in the information stream display interface of the client served by the first edge node server; and information stream exposure data representing the user's browsing state of loading information stream in the client served by the first edge node server. The client then sends the imperceptible loading data, the sliding distance residual distribution data, and the information stream exposure data to the first edge node server. The first edge node server is also used to send the seamless loading data, the sliding distance residual distribution data, and the information stream exposure data to the cloud server; The cloud server is used to determine whether the seamless loading data, the sliding distance residual distribution data, and the information stream exposure data are all acceptable data. If not, the cloud server adjusts the information stream loading timing data acquisition function until the seamless loading data, sliding distance residual distribution data, and information stream exposure data obtained based on the adjusted information stream loading timing data acquisition function are all acceptable data. The cloud server then sends the adjusted information stream loading timing data acquisition function to the first edge node server.

7. The system according to claim 1, characterized in that, The adjustment of the information stream loading method of the first edge node server includes: changing the compression method of the first edge node server for the view information in the information stream to be loaded from the first compression method to the second compression method, and sending a view information compression instruction for instructing the view information in the information stream to be loaded to be compressed according to the second compression method to the first edge node server; the compression loss data of the second compression method is greater than the compression loss data of the first compression method. The first edge node server is further configured to compress the view information in the information stream to be loaded according to the second compression method, obtain the compressed information stream to be loaded, and provide the compressed information stream to be loaded to the client.

8. The system according to claim 3, characterized in that, The step of determining the first client from among the clients served by the first edge node server that needs to adjust the information stream loading timing data includes: Obtain historical behavior data of the clients served by the edge node server during the execution of historical information stream loading tasks; Based on the historical behavior data, obtain information flow loading wait state data, which represents the user's waiting state when the user perceives the information flow loading process in the client served by the edge node server; Based on the information stream loading wait status data, determine the first client among the clients served by the first edge node server that needs to adjust the information stream loading timing data.

9. The system according to claim 2, characterized in that, The first edge node server is also used to acquire sliding speed fluctuation data representing the historical sliding speed fluctuation status of the user in the information flow display interface of the client; The step of obtaining information stream loading timing data, which indicates the timing of loading the information stream to be loaded in the client, based on the information stream browsing confidence data, the network quality index data, and the first load index data, includes: If the scrolling speed fluctuation data is less than a preset scrolling speed fluctuation threshold, then the first information stream loading timing data is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data, and is used as the information stream loading timing data; otherwise, the second information stream loading timing data is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data, and is used as the information stream loading timing data.

10. A method for processing data on the loading timing of an information stream, characterized in that, include: Obtain information stream browsing confidence data to represent the user's browsing of the information stream to be loaded in the client, wherein the information stream to be loaded is the information stream that has not been loaded in the client but needs to be loaded in the next information stream loading task; Obtain network quality indicator data used to represent the network quality of the client; Obtain first load metric data to represent the load status of a first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client; Based on the information stream browsing confidence data, the network quality index data, and the first load index data, information stream loading timing data is obtained to indicate the timing of loading the information stream to be loaded in the client.

11. A method for processing data on the loading timing of an information stream, characterized in that, include: In response to detecting that the user stops swiping in the graphical user interface of the client, information stream browsing confidence data is obtained to represent the user browsing the information stream to be loaded in the client, wherein the information stream to be loaded is the information stream that is not loaded in the client but needs to be loaded in the next information stream loading task; Obtain network quality indicator data to represent the network quality of the client; Send the information stream browsing confidence data and the network quality index data to the first edge node server, and receive information stream loading timing data returned by the first edge node server, which is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data used to indicate the load status of the first edge node server, to indicate the timing of loading the information stream to be loaded in the client.

12. A method for processing data on the loading timing of an information stream, characterized in that, include: Determine whether the triggering conditions for adjusting the loading timing of information stream data for the clients served by the first edge node server are met; If so, the information flow loading timing data of at least one client served by the first edge node server is adjusted so that the loading timing adjustment objective function value is an acceptable loading timing adjustment objective function value. The loading timing adjustment objective function includes a seamless loading parameter, an information flow usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception state of loading an information flow in a client served by the first edge node server. The information flow usage parameter represents the user's usage state of the information flow loaded in the client served by the first edge node server. The resource consumption parameter represents the resource consumption state of the first edge node server when loading an information flow in a client served by the first edge node server. The values ​​of the seamless loading parameter and the information flow usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. Send adjusted information stream loading timing data for at least one client served by the first edge node server to the first edge node server.

13. A data processing device for loading a first information stream, characterized in that, The device includes: The first acquisition unit is used to acquire information stream browsing confidence data representing the user's browsing of the information stream to be loaded in the client, wherein the information stream to be loaded is the information stream that is not loaded in the client but needs to be loaded in the next information stream loading task. The second acquisition unit is used to acquire network quality indicator data that represents the network quality of the client. The third acquisition unit is used to acquire first load index data for representing the load status of the first edge node server, wherein the first edge node server is a server that provides the information stream to be loaded to the client. The first obtaining unit is configured to obtain information flow loading timing data, which indicates the timing of loading the information flow to be loaded in the client, based on the information flow browsing confidence data, the network quality index data, and the first load index data.

14. A data processing device for second information stream loading timing, characterized in that, The device includes: The second obtaining unit is configured to obtain information stream browsing confidence data in response to detecting that the user stops swiping in the graphical user interface of the client, which represents the user browsing the information stream to be loaded in the client, wherein the information stream to be loaded is the information stream that is not loaded in the client but needs to be loaded in the next information stream loading task; The third obtaining unit is used to obtain network quality index data that represents the network quality of the client. The receiving unit is configured to send the information stream browsing confidence data and the network quality index data to the first edge node server, and receive information stream loading timing data returned by the first edge node server, which is obtained based on the information stream browsing confidence data, the network quality index data, and the first load index data used to indicate the load status of the first edge node server, and is used to indicate the timing of loading the information stream to be loaded in the client.

15. A data processing device for loading a third information stream, characterized in that, The device includes: The judgment unit is used to determine whether the triggering condition for adjusting the loading timing data of the information stream of the client served by the first edge node server is met. An adjustment unit is configured to adjust the information flow loading timing data of at least one client served by the first edge node server when a trigger condition for adjusting the information flow loading timing data of the client served by the first edge node server is met, so that the value of the loading timing adjustment objective function is an acceptable loading timing adjustment objective function value. The loading timing adjustment objective function includes a seamless loading parameter, an information flow usage parameter, and a resource consumption parameter. The seamless loading parameter represents the user's perception state of loading an information flow in the client served by the first edge node server. The information flow usage parameter represents the user's usage state of the information flow loaded in the client served by the first edge node server. The resource consumption parameter represents the resource consumption state of the first edge node server when loading an information flow in the client served by the first edge node server. The values ​​of the seamless loading parameter and the information flow usage parameter are positively correlated with the value of the loading timing adjustment objective function, and the value of the resource consumption parameter is negatively correlated with the value of the loading timing adjustment objective function. The sending unit is used to send the adjusted information stream loading timing data of at least one client served by the first edge node server to the first edge node server.

16. An electronic device, characterized in that, include: processor; as well as A memory for storing a data processing program, which, when the electronic device is powered on and runs through the processor, performs the method as described in any one of claims 10-12.

17. A computer-readable storage medium, characterized in that, The system contains a data processing program that is executed by a processor to perform the method as described in any one of claims 10-12.

Images