Self-construction network construction method and system

Abstract

The invention relates to the technical field of network construction, in particular to a self-construction network construction method and system. The method comprises the following steps: acquiring a network building demand; analyzing a network building demand, and determining a network use scene and a network use scale; determining time attribute characteristics and event attribute characteristics according to the network use scene and the network use scale; determining the response priority of the front end and the interface concurrency logic of the application layer according to the event attribute characteristics; and according to the time attribute characteristics, determining the response priority of the front end and the triggering opportunity of the interface concurrency logic of the application layer. An application environment and a scale index of network deployment are determined, demand ambiguity is eliminated, and optimization for a specific scene and scale is ensured. The content loading sequence and the interface calling mode are dynamically adjusted, so that the response efficiency is improved, and the system pressure is relieved. Configuration is dynamically started before or during the peak period of network flow, it is ensured that resource optimization takes effect in time, and invalid or delayed response is avoided.

Description

Technical Field

[0001] The present application relates to the technical field of network construction, and in particular to a method and system for constructing a self-built network. Background Art

[0002] With the rapid development of cloud computing technology, more and more companies are choosing to lease server resources from cloud service providers to build customized networks to meet their business needs. In existing technologies, cloud service providers typically design static network solutions based on the network requirements provided by their customers.

[0003] However, in high-concurrency scenarios such as user platforms, existing network construction solutions usually rely on fixed parameter configurations and cannot adapt to fluctuations in network usage, resulting in rigid resource allocation and low response efficiency, seriously affecting user experience and system stability. Summary of the Invention

[0004] This application provides a self-built network construction method and system to solve the above problems.

[0005] In a first aspect, the present application provides a method for establishing a self-built network, the method comprising: Obtain network construction requirements; analyze the network construction requirements and determine the network usage scenarios and network usage scale; Determining time attribute characteristics and event attribute characteristics based on the network usage scenario and the network usage scale; Determine the front-end response priority and application layer interface concurrency logic based on the event attribute characteristics; According to the time attribute characteristics, the response priority of the front end and the triggering timing of the interface concurrency logic of the application layer are determined.

[0006] Through this solution, the network construction requirements are obtained, and the analysis process is initiated according to the actual needs of the customer to avoid process interruptions due to missing requirements. The network construction requirements are analyzed to determine the network usage scenarios and network usage scale, clarify the application environment and scale indicators of network deployment, eliminate demand ambiguity, and ensure optimization for specific scenarios and scales. According to the network usage scenarios and network usage scale, the time attribute characteristics and event attribute characteristics are determined to effectively identify the dynamic changes of the network environment in the time and event dimensions. According to the event attribute characteristics, the front-end response priority and the interface concurrency logic of the application layer are determined, and the content loading order and interface calling method are dynamically adjusted to improve response efficiency and reduce system pressure. According to the time attribute characteristics, the front-end response priority and the triggering time of the interface concurrency logic of the application layer are determined, so as to dynamically enable the configuration before or during the network traffic peak, ensuring that resource optimization takes effect in a timely manner and avoiding invalid or delayed responses.

[0007] Optionally, determining the time attribute characteristics and event attribute characteristics according to the network usage scenario and the network usage scale includes: Acquire historical event data, analyze the historical event data, and determine event type classification rules; Analyzing the network usage scenario based on the event type classification rule to determine the type of network congestion event; Based on the type of network congestion event, analyze the historical event data to determine historical traffic peak characteristics and interface call path characteristics; According to the historical traffic peak characteristics and the interface call path characteristics, the time attribute characteristics and the event attribute characteristics are determined.

[0008] Through this solution, historical event data is obtained and analyzed, and event type classification rules are determined to ensure that network congestion event types can be accurately identified, avoiding the problem of lack of dynamic analysis capabilities. Based on the event type classification rules, network usage scenarios are analyzed to determine the type of network congestion events, solving the problem of unutilized event attribute characteristics. Based on the network congestion event type, historical event data is analyzed to determine the historical traffic peak characteristics and interface call path characteristics, quantify the dynamic behavior under the event type, provide a data basis for attribute characteristic generation, and eliminate the defects of missing predictions of time attribute characteristics and event attribute characteristics. Based on the historical traffic peak characteristics and interface call path characteristics, the time attribute characteristics and event attribute characteristics are determined to achieve the goal of dynamically determining time attribute and event attribute characteristics and improve response efficiency.

[0009] Optionally, determining the front-end response priority and the interface concurrency logic of the application layer according to the event attribute characteristics includes: Determine the user's browsing focus and browsing page content based on the event attribute characteristics; Analyze the historical event data to determine event browsing preferences; Splitting the browsing page content to obtain a plurality of content units; Based on the event browsing preference, a plurality of content units are prioritized to obtain a response priority of the front end; Generate application layer interface concurrency logic based on the historical traffic peak characteristics and the interface call path characteristics.

[0010] Through this solution, the user's browsing core and browsing page content are determined according to the characteristics of event attributes, avoiding resource waste on low-relevance pages, while ensuring that optimization actions are focused on the actual source of congestion. Analyze historical event data, determine event browsing preferences, and ensure that response priorities fit real user behavior rather than static preset rules. Split the browsing page content to obtain several content units, so that optimization can be accurate to the smallest functional module, avoiding resource redundancy for full page loading. Based on event browsing preferences, prioritize several content units to obtain the front-end response priority, significantly shortening user-perceived waiting time and alleviating page freeze issues. Based on historical traffic peak characteristics and interface call path characteristics, generate application layer interface concurrency logic, dynamically disperse concurrency pressure, and eliminate the defect of interface response delay.

[0011] Optionally, determining the time attribute characteristics and the event attribute characteristics based on the historical traffic peak characteristics and the interface call path characteristics includes: Analyze the historical traffic peak characteristics to determine periodic time patterns and sudden time patterns; Analyze the interface call path characteristics to determine the interface response delay pattern and interface error rate pattern; Time attribute characteristics and event attribute characteristics are determined according to the periodic time pattern, the burst time pattern, the interface response delay pattern, and the interface error rate pattern.

[0012] This solution analyzes historical traffic peak characteristics to identify periodic and sudden time patterns, addressing the lack of dynamic analysis capabilities. Historical traffic peak characteristics generate predictable patterns to avoid resource optimization delays. Interface call path characteristics are analyzed to identify interface response delay patterns and interface error rate patterns, addressing the lack of response priority and concurrency logic. Pattern data is generated from interface call path characteristics to ensure adaptive settings for interface concurrency logic. Based on periodic and sudden time patterns, interface response delay patterns, and interface error rate patterns, time attribute characteristics and event attribute characteristics are determined to address insufficient scalability. By integrating pattern data to generate attribute characteristics, dynamic configuration adjustments can be made as network usage scales.

[0013] Optionally, the historical event data includes historical traffic data, and the analyzing the network usage scenario based on the event type classification rule to determine the type of network congestion event includes: Extracting the network usage scenario and determining scenario characteristic parameters; Analyze the historical traffic data and identify traffic anomalies; Based on the event type classification rule, according to the traffic anomaly point and the scene feature parameters, determine the initial matching type; The initial matching type is verified in combination with the network usage scale to determine the type of network congestion event.

[0014] This solution extracts network usage scenarios and identifies scenario-specific parameters, addressing the lack of event attribute characteristics. It analyzes historical traffic data to identify traffic anomalies and expose unmanageable congestion bursts. Based on event type classification rules, it determines the initial matching type based on traffic anomalies and scenario-specific parameters, addressing the inability to predict network congestion event types and avoiding the blindness of static solutions. The initial matching type is verified based on network usage scale to determine the type of network congestion event, ensuring that the results meet actual scale requirements and avoiding resource misallocation.

[0015] Optionally, analyzing the historical event data to determine event browsing preferences includes: Parsing the historical event data to obtain user click stream data; Analyze the user click stream data to determine page dwell time distribution and click heat map; Determining a content attention index based on the page dwell time distribution and the click heat map; Combined with the event attribute characteristics and the content attention index, low-attention events are filtered to determine event browsing preferences.

[0016] This solution parses historical event data to generate user clickstream data, addressing the lack of dynamic analysis capabilities. By analyzing user clickstream data, we determine page dwell time distribution and click heatmaps, eliminating issues related to missing response priorities and concurrency logic. Based on page dwell time distribution and click heatmaps, we determine content attention metrics, ensuring that event browsing preferences are based on quantitative data and addressing scalability issues. By combining event attribute characteristics with content attention metrics, we filter low-attention events, determine event browsing preferences, and address issues related to missing response priorities and concurrency logic.

[0017] Optionally, prioritizing a plurality of content units based on the event browsing preference to obtain a front-end response priority includes: assigning a preference weight to each content unit based on the event browsing preference; Analyze the historical traffic peak characteristics and determine the unit loading delay tolerance; Calculating a priority score based on the preference weight and the unit loading delay tolerance; Based on the priority scores, the plurality of content units are sorted to obtain a response priority of the front end.

[0018] Through this solution, preference weights are assigned to each content unit based on event browsing preferences, ensuring that response priority sorting can give priority to high-preference content units, thereby optimizing the initial basis for front-end loading in a dynamic network environment. Analyze historical traffic peak characteristics, determine the unit loading delay tolerance, identify the performance weaknesses of each content unit under stress scenarios, and ensure that response priority sorting can give priority to low-tolerance content units, thereby alleviating interface congestion. Calculate priority scores based on preference weights and unit loading delay tolerance to ensure response priority while taking into account user behavior and network performance requirements. Based on the priority scores, sort several content units to obtain the front-end response priority, ensuring that high-priority content units respond quickly in dynamic events, reducing delays and error rates, optimizing content loading order, and solving the problem of missing response priorities.

[0019] Optionally, generating application layer interface concurrency logic according to the historical traffic peak characteristics and the interface call path characteristics includes: Analyze the historical traffic peak characteristics and determine the peak concurrency threshold; Analyze the interface call path characteristics and determine the interface dependency graph; Generate an interface call concurrency strategy and a timeout fallback mechanism based on the peak concurrency threshold and the interface dependency graph; The interface call concurrency strategy and the timeout fallback mechanism are used as the interface concurrency logic of the application layer.

[0020] Through this solution, the historical traffic peak characteristics are analyzed, the peak concurrency threshold is determined, the maximum concurrency pressure of the historical peak event is reflected, and the defect of being unable to predict the time attribute characteristics is resolved. The interface call path characteristics are analyzed, the interface dependency graph is determined, and the defect of being unable to parse the event attribute characteristics is eliminated. According to the peak concurrency threshold and the interface dependency graph, the interface call concurrency strategy and timeout fallback mechanism are generated to optimize the calling logic of the application layer interface, prevent congestion caused by multiple triggers of the same interface at the same time, and alleviate the pressure in high concurrency scenarios; at the same time, the response timeout or error problem is eliminated, and the entire link is avoided from being paralyzed due to the blocking of a single interface, reducing the deterioration of user experience. The interface call concurrency strategy and timeout fallback mechanism are used as the interface concurrency logic of the application layer to eliminate the defect of missing application layer interface concurrency logic.

[0021] Optionally, analyzing the historical event data to determine event type classification rules includes: Parsing the historical event data to obtain user behavior logs and network traffic logs; Analyze the user behavior log to determine user behavior patterns; Analyzing the network traffic logs to determine traffic fluctuation patterns; An event type classification rule is determined based on the user behavior pattern and the traffic fluctuation pattern.

[0022] This solution analyzes historical event data to generate user behavior logs and network traffic logs. This allows for correlation between user behavior pattern analysis, supports quantification of traffic fluctuation patterns, and avoids coupling operation records with traffic metrics during analysis. By analyzing user behavior logs and identifying user behavior patterns, we address the inability to predict event attributes. By analyzing network traffic logs and identifying traffic fluctuation patterns, we address the inability to predict temporal attributes. Based on user behavior and traffic fluctuation patterns, we determine event classification rules, eliminating the core drawback of a lack of dynamic analysis capabilities.

[0023] In a second aspect, the present application provides a self-built network construction system, the system comprising: A demand analysis module is used to obtain network construction requirements; analyze the network construction requirements, and determine the network usage scenario and network usage scale; a characteristic determination module, configured to determine time attribute characteristics and event attribute characteristics according to the network usage scenario and the network usage scale; An event characteristics analysis module is used to determine the front-end response priority and the interface concurrency logic of the application layer based on the event attribute characteristics; The time characteristic analysis module is used to determine the response priority of the front end and the triggering time of the interface concurrency logic of the application layer according to the time attribute characteristics.

[0024] Optionally, when the characteristic determination module determines the time attribute characteristics and the event attribute characteristics according to the network usage scenario and the network usage scale, it is configured to: Acquire historical event data, analyze the historical event data, and determine event type classification rules; Analyzing the network usage scenario based on the event type classification rule to determine the type of network congestion event; Based on the type of network congestion event, analyze the historical event data to determine historical traffic peak characteristics and interface call path characteristics; According to the historical traffic peak characteristics and the interface call path characteristics, the time attribute characteristics and the event attribute characteristics are determined.

[0025] Optionally, when the event characteristic analysis module determines the response priority of the front end and the interface concurrency logic of the application layer based on the event attribute characteristics, it is used to: Determine the user's browsing focus and browsing page content based on the event attribute characteristics; Analyze the historical event data to determine event browsing preferences; Splitting the browsing page content to obtain a plurality of content units; Based on the event browsing preference, a plurality of content units are prioritized to obtain a response priority of the front end; Generate application layer interface concurrency logic based on the historical traffic peak characteristics and the interface call path characteristics.

[0026] Optionally, when the characteristic determination module determines the time attribute characteristics and the event attribute characteristics based on the historical traffic peak characteristics and the interface call path characteristics, it is used to: Analyze the historical traffic peak characteristics to determine periodic time patterns and sudden time patterns; Analyze the interface call path characteristics to determine the interface response delay pattern and interface error rate pattern; Time attribute characteristics and event attribute characteristics are determined according to the periodic time pattern, the burst time pattern, the interface response delay pattern, and the interface error rate pattern.

[0027] Optionally, the historical event data includes historical traffic data, and the characteristic determination module analyzes the network usage scenario based on the event type classification rule to determine the type of network congestion event, and is configured to: Extracting the network usage scenario and determining scenario characteristic parameters; Analyze the historical traffic data and identify traffic anomalies; Based on the event type classification rule, according to the traffic anomaly point and the scene feature parameters, determine the initial matching type; The initial matching type is verified in combination with the network usage scale to determine the type of network congestion event.

[0028] Optionally, when the event feature analysis module analyzes the historical event data and determines the event browsing preference, it is used to: Parsing the historical event data to obtain user click stream data; Analyze the user click stream data to determine page dwell time distribution and click heat map; Determining a content attention index based on the page dwell time distribution and the click heat map; Combined with the event attribute characteristics and the content attention index, low-attention events are filtered to determine event browsing preferences.

[0029] Optionally, the event feature analysis module prioritizes a plurality of content units based on the event browsing preference, and when obtaining the front-end response priority, is used to: assigning a preference weight to each content unit based on the event browsing preference; Analyze the historical traffic peak characteristics and determine the unit loading delay tolerance; Calculating a priority score based on the preference weight and the unit loading delay tolerance; Based on the priority scores, the plurality of content units are sorted to obtain a response priority of the front end.

[0030] Optionally, when the event characteristic analysis module generates the interface concurrency logic of the application layer based on the historical traffic peak characteristics and the interface call path characteristics, it is used to: Analyze the historical traffic peak characteristics and determine the peak concurrency threshold; Analyze the interface call path characteristics and determine the interface dependency graph; Generate an interface call concurrency strategy and a timeout fallback mechanism based on the peak concurrency threshold and the interface dependency graph; The interface call concurrency strategy and the timeout fallback mechanism are used as the interface concurrency logic of the application layer.

[0031] Optionally, when the feature determination module analyzes the historical event data and determines the event type classification rules, it is used to: Parsing the historical event data to obtain user behavior logs and network traffic logs; Analyze the user behavior log to determine user behavior patterns; Analyzing the network traffic logs to determine traffic fluctuation patterns; An event type classification rule is determined based on the user behavior pattern and the traffic fluctuation pattern. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0033] Figure 1 A schematic diagram of an application scenario provided in one embodiment of the present application; Figure 2 A flowchart of a method for establishing a self-built network according to an embodiment of the present application; Figure 3 A schematic diagram of the structure of a self-built network construction system provided in one embodiment of the present application. DETAILED DESCRIPTION

[0034] To make the purpose, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0035] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates an "or" relationship between the related objects.

[0036] The embodiments of the present application are described in further detail below with reference to the accompanying drawings.

[0037] In high-concurrency scenarios like user platforms, existing network deployment solutions often rely on fixed parameter configurations, resulting in rigid resource allocation and inefficient response, severely impacting user experience and system stability. In particular, in specific event-driven scenarios, they are unable to adapt to fluctuations in network usage in real time, leading to degraded system performance and a worsened user experience.

[0038] Based on this, the present application provides a self-built network construction method and system to obtain network construction requirements, start the parsing process according to the actual needs of the customer, and avoid process interruptions due to missing requirements. Analyze network construction requirements, determine network usage scenarios and network usage scales, clarify the application environment and scale indicators of network deployment, eliminate demand ambiguity, and ensure optimization for specific scenarios and scales. According to the network usage scenarios and network usage scales, determine the time attribute characteristics and event attribute characteristics, and effectively identify the dynamic changes of the network environment in the time and event dimensions. According to the event attribute characteristics, determine the front-end response priority and the interface concurrency logic of the application layer, and dynamically adjust the content loading order and interface calling method, so as to improve response efficiency and reduce system pressure. According to the time attribute characteristics, determine the front-end response priority and the triggering time of the interface concurrency logic of the application layer, so as to dynamically enable the configuration before or during the network traffic peak, to ensure that resource optimization takes effect in time, and avoid invalid or delayed responses.

[0039] Figure 1 This is a schematic diagram of an application scenario provided by this application. When setting up a self-built network, the method provided by this application is applied.

[0040] Specifically, the method provided in the present application is applied to any server, and the server interacts with the user device, obtains the network construction requirements submitted by the customer through the user device, and starts the parsing process according to the actual needs of the customer. Parse the network construction requirements and determine the network usage scenario and network usage scale. According to the network usage scenario and the network usage scale, determine the time attribute characteristics and event attribute characteristics. According to the event attribute characteristics, determine the response priority of the front end and the interface concurrency logic of the application layer, and dynamically adjust the content loading order and interface calling method to improve response efficiency and reduce system pressure. According to the time attribute characteristics, determine the response priority of the front end and the triggering time of the interface concurrency logic of the application layer, so as to dynamically enable the configuration before or during the network traffic peak, to ensure that resource optimization takes effect in time and avoid invalid or delayed responses.

[0041] For specific implementation methods, please refer to the following embodiments.

[0042] Figure 2 This is a flow chart of a self-built network construction method provided in an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201. Obtain network construction requirements; analyze network construction requirements, and determine network usage scenarios and network usage scale; Network construction requirements can be the demand information submitted by customers when leasing cloud services, including parameters such as bandwidth requirements, capacity requirements, and latency requirements. Network usage scenarios can be the application environment type of network deployment. Network usage scale can be a scale indicator of network deployment.

[0043] Specifically, the network construction requirements submitted by customers are obtained through user devices. A natural language processing engine is used to parse these requirements and match them with a list of keywords (for example, e-commerce keywords include promotions and shopping carts, and education keywords include courses and live streaming). This determines the network usage scenario (such as a user platform or an online education platform). Furthermore, numerical parameters (for example, bandwidth: 100 in 100Mbps) are extracted from the network construction requirements to determine the scale of network usage.

[0044] S202: Determine time attribute characteristics and event attribute characteristics based on the network usage scenario and network usage scale; The time attribute feature can be a characteristic pattern of traffic peak in the time dimension.

[0045] Event attribute characteristics can be event features driven by user behavior patterns.

[0046] Specifically, the historical event database (which stores historical traffic data and user behavior logs) established by the user platform's network monitoring system is matched with records of similar network usage scenarios and network usage scales to extract historical traffic data (for example, for e-commerce scenarios, traffic logs during festivals are retrieved); statistical analysis tools are used to process historical traffic data to identify temporal attribute characteristics (time series characteristics, such as the periodic pattern of traffic peaks (peaks at fixed times of the day) and the timing of sudden event triggering (instantaneous traffic surges caused by festival activities)).

[0047] At the same time, we analyze user behavior logs in the historical event database (such as clickstream data) and extract event attribute characteristics (such as high concurrency triggering interfaces in flash sale events and browsing preferences leading to interface congestion).

[0048] S203. Determine the front-end response priority and the application layer interface concurrency logic based on the event attribute characteristics; The front end can be the user interface, involving page rendering and content display, and processing responses to user requests. Response priority can be the order in which content units are loaded. The application layer can be the logic processing layer in the software architecture, involving interface calls and business logic execution, used to process application layer requests. Interface concurrency logic can be the interface call strategy.

[0049] Specifically, the front-end response priority is set according to the characteristics of the event attributes. For example, in an e-commerce scenario, if the event attribute characteristics show that the user prefers gift pages, the response priority is set to load this type of content first (high-priority content is processed first).

[0050] Set the application layer interface concurrency logic based on the characteristics of the event attributes. For example, for high-concurrency events (operations with a large number of requests in a short period of time, such as flash sales), increase the thread pool size (indicates the number of threads that can process requests simultaneously and adjust concurrency control parameters, such as the number of threads).

[0051] S204: Determine the response priority of the front end and the triggering timing of the interface concurrency logic of the application layer according to the time attribute characteristics.

[0052] The triggering timing may be the activation time point of the response priority and interface concurrency logic.

[0053] Specifically, analyze the time attribute characteristics, combine them with event triggers (input time attribute characteristics to determine the trigger timing in real time), and generate the front-end response priority and the trigger timing of the application layer interface concurrency logic.

[0054] Through this solution, the network construction requirements are obtained, and the analysis process is initiated according to the actual needs of the customer to avoid process interruptions due to missing requirements. The network construction requirements are analyzed to determine the network usage scenarios and network usage scale, clarify the application environment and scale indicators of network deployment, eliminate demand ambiguity, and ensure optimization for specific scenarios and scales. According to the network usage scenarios and network usage scale, the time attribute characteristics and event attribute characteristics are determined to effectively identify the dynamic changes of the network environment in the time and event dimensions. According to the event attribute characteristics, the front-end response priority and the interface concurrency logic of the application layer are determined, and the content loading order and interface calling method are dynamically adjusted to improve response efficiency and reduce system pressure. According to the time attribute characteristics, the front-end response priority and the triggering time of the interface concurrency logic of the application layer are determined, so as to dynamically enable the configuration before or during the network traffic peak, ensuring that resource optimization takes effect in a timely manner and avoiding invalid or delayed responses.

[0055] In some embodiments, historical event data is obtained, analyzed, and event type classification rules are determined; based on the event type classification rules, network usage scenarios are analyzed to determine network congestion event types; based on network congestion event types, historical event data is analyzed to determine historical traffic peak characteristics and interface call path characteristics; based on historical traffic peak characteristics and interface call path characteristics, time attribute characteristics and event attribute characteristics are determined.

[0056] Historical event data can be a historical record dataset related to network usage scenarios, including historical traffic data and user behavior logs. Event type classification rules can be logical conditions used to classify network usage scenarios into different event types. Network congestion event types can be the congestion event categories determined based on event type classification rules. Historical traffic peak features can be quantitative patterns related to traffic surges extracted from historical event data. Interface call path features can be behavioral patterns related to application layer interface calls extracted from historical event data.

[0057] Specifically, historical event data is extracted from the historical event database collected through the supplier's log server; statistical analysis tools are used to process historical event data, identify event characteristics (such as event frequency, user behavior patterns), and then formulate event type classification rules based on event characteristics. For example, events are divided into high-concurrency events or low-concurrency events according to user concurrency.

[0058] Apply event type classification rules to the current network usage scenario. By matching scenario features (such as descriptions of flash sales on the user platform), identify the types of network congestion events that may cause network congestion. For example, in an e-commerce scenario, the rules match flash sales or holiday promotions as network congestion event types.

[0059] For each type of network congestion event (such as flash sales), filter historical event data, use statistical analysis tools to analyze the data, calculate the peak traffic time point (peak occurrence time) and peak amplitude (maximum request volume), and thus determine the historical traffic peak characteristics; and identify frequently called interface sequences (such as the interface call chain from users browsing the page to placing an order) to determine the interface call path characteristics.

[0060] Analyze the time point data in historical traffic peak characteristics to generate time attribute characteristics, such as the daily periodic occurrence of traffic peaks. Analyze the interface call path characteristics to generate event attribute characteristics, such as the user preference for calling the gift page interface in a flash sale event.

[0061] Through this solution, historical event data is obtained and analyzed, and event type classification rules are determined to ensure that network congestion event types can be accurately identified, avoiding the problem of lack of dynamic analysis capabilities. Based on the event type classification rules, network usage scenarios are analyzed to determine the type of network congestion events, solving the problem of unutilized event attribute characteristics. Based on the network congestion event type, historical event data is analyzed to determine the historical traffic peak characteristics and interface call path characteristics, quantify the dynamic behavior under the event type, provide a data basis for attribute characteristic generation, and eliminate the defects of missing predictions of time attribute characteristics and event attribute characteristics. Based on the historical traffic peak characteristics and interface call path characteristics, the time attribute characteristics and event attribute characteristics are determined to achieve the goal of dynamically determining time attribute and event attribute characteristics and improve response efficiency.

[0062] In some embodiments, based on the characteristics of event attributes, the user's browsing core and browsing page content are determined; historical event data is analyzed to determine event browsing preferences; the browsing page content is split to obtain several content units; based on the event browsing preferences, the several content units are prioritized to obtain the front-end response priority; based on the historical traffic peak characteristics and interface call path characteristics, the application layer interface concurrency logic is generated.

[0063] The user browsing core can be the core pages or functional interfaces that users visit most frequently.

[0064] The browsing page content can be a complete set of pages associated with the user's browsing core in the network usage scenario.

[0065] Event browsing preference can be a quantitative indicator of the frequency with which users access different browsing page contents or interfaces.

[0066] A content unit is the smallest independently loadable component of a page's content.

[0067] Specifically, analyze user behavior patterns in event attribute characteristics (such as high-frequency calls to gift interfaces or user concurrency patterns) and identify the user's browsing focus. For example, in a flash sale event on a user platform, the event attribute characteristics include the user's preference to call the gift page interface, and based on this, determine that the user's browsing focus is the gift page.

[0068] Based on the user's browsing focus, query the customer description in the network construction requirements (such as the project requirements submitted by the customer) and extract the corresponding browsing page content. For example, extract the gift page, product details page, and shopping cart page from the customer description as the browsing page content.

[0069] Use statistical analysis tools to process historical event data and filter the data to match the type of network congestion event (such as flash sale events). Analyze user behavior logs in historical event data (such as page browsing records) to determine event browsing preferences. For example, in flash sale events, analyze historical event data to determine the frequency of visits to the gift page as the event browsing preference.

[0070] Use page parsing tools to break down the browsing page content into several content units (e.g., page title, product image, description text, and button). For example, a gift page is broken down into a gift image unit, a gift price unit, and a buy button unit. Based on event browsing preferences, each content unit is assigned a weight based on the preference value of the page to which it belongs. Multiple content units on the same page are then ranked by interface call frequency in user behavior logs (e.g., the buy button unit has the highest call frequency on the gift page). A sorting algorithm is then used to sort all content units in descending order of weight to generate a front-end response priority, e.g., the buy button unit, gift image unit, and gift price unit. Peak time windows are extracted from historical traffic peak characteristics to generate time-driven rules (used to control interface concurrency within different time windows and generate application-layer interface concurrency logic). Highly concurrent interfaces are extracted from interface call path characteristics to generate interface scheduling rules (used to optimize interface scheduling strategies and generate application-layer interface concurrency logic). Time-driven rules and interface scheduling rules are combined to generate application-layer interface concurrency logic.

[0071] Through this solution, the user's browsing core and browsing page content are determined according to the characteristics of event attributes, avoiding resource waste on low-relevance pages, while ensuring that optimization actions are focused on the actual source of congestion. Analyze historical event data, determine event browsing preferences, and ensure that response priorities fit real user behavior rather than static preset rules. Split the browsing page content to obtain several content units, so that optimization can be accurate to the smallest functional module, avoiding resource redundancy for full page loading. Based on event browsing preferences, prioritize several content units to obtain the front-end response priority, significantly shortening user-perceived waiting time and alleviating page freeze issues. Based on historical traffic peak characteristics and interface call path characteristics, generate application layer interface concurrency logic, dynamically disperse concurrency pressure, and eliminate the defect of interface response delay.

[0072] In some embodiments, historical traffic peak characteristics are analyzed to determine periodic time patterns and burst time patterns; interface call path characteristics are analyzed to determine interface response delay patterns and interface error rate patterns; and time attribute characteristics and event attribute characteristics are determined based on periodic time patterns, burst time patterns, interface response delay patterns, and interface error rate patterns.

[0073] A periodic time pattern can be a fixed time pattern that recurs repeatedly.

[0074] A bursty time pattern can be an abnormal traffic peak with no fixed pattern.

[0075] The interface response delay pattern may be a response delay distribution rule of the interface.

[0076] The interface error rate pattern may be a probability law of interface call failure.

[0077] Specifically, statistical analysis tools are used to detect repetitive cycles in historical traffic peak characteristics (such as daily or weekly peak hours), calculate the traffic average value (used to identify periodic time patterns) and standard deviation (used to quantify traffic fluctuations and assist in identifying periodic time patterns) within the time window, identify traffic peaks at fixed intervals (high point values ​​at fixed intervals in historical traffic peak characteristics, used to determine periodic time patterns), and determine periodic time patterns; apply anomaly detection algorithms to calculate the deviation between the traffic value (used to detect deviations) and the historical average value (long-term average traffic value, used to detect deviations), and identify sudden peaks (abnormal traffic high points that deviate significantly from the historical average value, used to determine sudden time patterns), and determine sudden time patterns.

[0078] Analyze the interface call path characteristics and calculate the average response time of each interface (the arithmetic mean of the interface response time, used to determine the interface response delay pattern); identify high-latency interfaces (interfaces with excessively long response times) by comparing the average response times and determine the interface response delay pattern; at the same time, analyze the interface call path characteristics and calculate the error rate of each interface (the proportion of interface call failures); identify high-error rate interfaces (interfaces with excessively high error rates) by comparing the error rates and determine the interface error rate pattern.

[0079] Map periodic time patterns into the periodic portion of a time attribute characteristic (describing the regular recurrence of traffic peaks, such as daily or weekly peaks); map sudden time patterns into the sudden portion of a time attribute characteristic (describing the abnormal and irregular surges of traffic peaks, such as promotional events); and combine the periodic and sudden portions to generate a time attribute characteristic. Map interface response delay patterns into the response delay characteristics of an event attribute characteristic (describing the interface delay performance characteristics, such as the distribution of high-latency interfaces); map interface error rate patterns into the error rate characteristics of an event attribute characteristic (describing the interface error rate performance characteristics, such as the distribution of high-error rate interfaces); and combine the response delay characteristics and error rate characteristics to generate an event attribute characteristic.

[0080] This solution analyzes historical traffic peak characteristics to identify periodic and sudden time patterns, addressing the lack of dynamic analysis capabilities. Historical traffic peak characteristics generate predictable patterns to avoid resource optimization delays. Interface call path characteristics are analyzed to identify interface response delay patterns and interface error rate patterns, addressing the lack of response priority and concurrency logic. Pattern data is generated from interface call path characteristics to ensure adaptive settings for interface concurrency logic. Based on periodic and sudden time patterns, interface response delay patterns, and interface error rate patterns, time attribute characteristics and event attribute characteristics are determined to address insufficient scalability. By integrating pattern data to generate attribute characteristics, dynamic configuration adjustments can be made as network usage scales.

[0081] In some embodiments, network usage scenarios are extracted and scenario characteristic parameters are determined; historical traffic data is analyzed to identify traffic anomalies; based on event type classification rules, the initial matching type is determined according to traffic anomalies and scenario characteristic parameters; combined with the network usage scale, the initial matching type is verified and the type of network congestion event is determined.

[0082] The scenario characteristic parameters may be quantitative characteristic parameters extracted from the network usage scenario and used to describe the dynamic characteristics of the application environment.

[0083] Historical traffic data can be used to analyze the scale and time attributes of network usage in historical event data and is a traffic-related indicator.

[0084] Traffic anomalies may be abnormal fluctuation points identified when analyzing historical traffic data.

[0085] The initial matching type may be a network congestion event type preliminarily determined according to an event type classification rule.

[0086] Specifically, network usage scenarios (such as user platforms or online education platforms) are extracted from the network construction requirements submitted by customers. Based on the network usage scenarios, the preset scenario database constructed through historical event data is queried (storing common scenario templates in historical event data) to map out scenario feature parameters. For example, for the user platform, the scenario feature parameters include page types including product details page and checkout page, and event triggering frequency including high-concurrency interface call mode.

[0087] Calculate the average and standard deviation of the historical traffic data in the specified time window (a fixed time range used to calculate the average and standard deviation of the traffic data) in the historical event data; compare the current traffic value with the historical average value. If the comparison result exceeds the standard deviation threshold set based on the historical event data (used to determine whether the traffic value is abnormal), it is marked as a traffic anomaly point.

[0088] Based on the traffic anomalies and scene characteristic parameters, the similarity score is calculated using the event type classification rules. Based on the similarity score, the initial matching type is determined. For example, if the scene characteristic parameters are e-commerce and the traffic anomaly occurs during a holiday, the match is a holiday promotion event.

[0089] The network usage scale is compared with the scale threshold set based on historical event data corresponding to the initial matching type (used to verify whether the network usage scale causes congestion). If the network usage scale exceeds the scale threshold, the verification passes and the network congestion event type is determined. If the network usage scale does not reach the scale threshold, the event type classification rules are adjusted (such as lowering the similarity score), the initial matching type is re-determined, and verification is repeated until the final network congestion event type is determined.

[0090] This solution extracts network usage scenarios and identifies scenario-specific parameters, addressing the lack of event attribute characteristics. It analyzes historical traffic data to identify traffic anomalies and expose unmanageable congestion bursts. Based on event type classification rules, it determines the initial matching type based on traffic anomalies and scenario-specific parameters, addressing the inability to predict network congestion event types and avoiding the blindness of static solutions. The initial matching type is verified based on network usage scale to determine the type of network congestion event, ensuring that the results meet actual scale requirements and avoiding resource misallocation.

[0091] In some embodiments, historical event data is parsed to obtain user click stream data; user click stream data is analyzed to determine page dwell time distribution and click heat map; content attention index is determined based on page dwell time distribution and click heat map; low-attention events are filtered out in combination with event attribute characteristics and content attention index to determine event browsing preferences.

[0092] User click stream data can be a record of the sequence of clicks made by users on a page.

[0093] Page dwell time distribution can be the statistical distribution of the length of time a user stays on a single page.

[0094] A click heatmap can be a density distribution map of the locations users click on a page.

[0095] The content attention index may be a numerical index that quantifies the degree of user attention to the content of a page.

[0096] A low-attention event may be an event in which the content attention index is lower than a preset threshold.

[0097] Specifically, parse the user behavior logs in the historical event data, identify the fields of each user behavior log record, such as user identifier (such as user ID), page access event (such as the page path visited), click event (such as button or link click), timestamp (record the exact time when the event occurred), etc.; aggregate these records to generate user click stream data.

[0098] Based on user clickstream data, for each page (such as the product details page on the user platform), calculate the user's dwell time on each page (the difference between the page entry timestamp and the page exit timestamp). Aggregate the dwell time of all users to generate a page dwell time distribution. For example, calculate the frequency distribution of the dwell time on the product details page and display the proportion of users in different dwell time periods.

[0099] Click events (records of users' clicks on interactive elements (such as buttons, links, or images) on different pages) are extracted from user clickstream data and mapped to page areas (such as the location of buttons and images on the page). The click frequency of each page area (such as the number of times the payment button is clicked on the checkout page) is counted and visualized as a click heat map (where high-density areas indicate high-frequency click locations).

[0100] Assign weights to the page dwell time distribution and click heat map (where long dwell time indicates high attention, and high click density indicates high attention); then calculate the weighted score of the page dwell time distribution weight and the click heat map weight as the content attention indicator.

[0101] Based on the characteristics of event attributes and content attention indicators, a content attention indicator is associated with each event (such as a single user session or page visit event); a preset threshold is pre-set based on statistical experience with historical event data (used as a judgment criterion when filtering low-attention events); all events are traversed, and events with content attention indicators lower than the preset threshold (i.e., low-attention events) are removed; and the page identifier frequency distribution of the remaining events is then counted to determine the event browsing preference.

[0102] This solution parses historical event data to generate user clickstream data, addressing the lack of dynamic analysis capabilities. By analyzing user clickstream data, we determine page dwell time distribution and click heatmaps, eliminating issues related to missing response priorities and concurrency logic. Based on page dwell time distribution and click heatmaps, we determine content attention metrics, ensuring that event browsing preferences are based on quantitative data and addressing scalability issues. By combining event attribute characteristics with content attention metrics, we filter low-attention events, determine event browsing preferences, and address issues related to missing response priorities and concurrency logic.

[0103] In some embodiments, a preference weight is assigned to each content unit based on event browsing preferences; historical traffic peak characteristics are analyzed to determine the unit loading delay tolerance; a priority score is calculated based on the preference weight and the unit loading delay tolerance; and based on the priority score, several content units are sorted to obtain the front-end response priority.

[0104] The preference weight may be a numerical value representing the strength of user preference for each content unit.

[0105] The unit loading delay tolerance may be a time threshold indicating the upper limit of the maximum delay acceptable during the loading process of each content unit.

[0106] The priority score may be a numerical value used to quantify the loading priority order of each content unit.

[0107] Specifically, all content units are traversed (for example, all content units are identified on the front-end page of the user platform, such as product images and purchase buttons); for each content unit, its corresponding frequency value in the event browsing preference is queried; and based on the frequency value, a preference weight is assigned to each content unit.

[0108] Analyze historical traffic peak characteristics and identify the performance of each content unit in historical peak events (instances with abnormally high traffic in historical event data (such as flash sales)) (such as extracting the average response delay of the unit during the peak period from user behavior logs); then determine the unit loading delay tolerance based on preset rules established based on historical performance data. For example, compare the performance of content units in historical peak events. If the average delay of the content unit during the peak period is higher, reduce its tolerance; if the average delay of the content unit during the peak period is lower, increase its tolerance.

[0109] A weighted formula is used to calculate a priority score based on the preference weight and unit loading delay tolerance. A high score indicates a high loading priority (high preference weight and low unit loading delay tolerance), while a low score indicates a low priority. Based on the priority score, a quick sort algorithm is used to sort the content units in descending order of priority score, generating a front-end response priority. The highest-scoring units are loaded first.

[0110] Through this solution, preference weights are assigned to each content unit based on event browsing preferences, ensuring that response priority sorting can give priority to high-preference content units, thereby optimizing the initial basis for front-end loading in a dynamic network environment. Analyze historical traffic peak characteristics, determine the unit loading delay tolerance, identify the performance weaknesses of each content unit under stress scenarios, and ensure that response priority sorting can give priority to low-tolerance content units, thereby alleviating interface congestion. Calculate priority scores based on preference weights and unit loading delay tolerance to ensure response priority while taking into account user behavior and network performance requirements. Based on the priority scores, sort several content units to obtain the front-end response priority, ensuring that high-priority content units respond quickly in dynamic events, reducing delays and error rates, optimizing content loading order, and solving the problem of missing response priorities.

[0111] In some embodiments, historical traffic peak characteristics are analyzed to determine the peak concurrency threshold; interface call path characteristics are analyzed to determine the interface dependency graph; based on the peak concurrency threshold and the interface dependency graph, an interface call concurrency strategy and a timeout fallback mechanism are generated; and the interface call concurrency strategy and the timeout fallback mechanism are used as the interface concurrency logic of the application layer.

[0112] The peak concurrency threshold may be the upper limit of the maximum number of concurrent calls allowed by the application layer interface.

[0113] The interface dependency graph may be a graph data structure representing the call dependency relationship between application layer interfaces.

[0114] The interface call concurrency policy may be a concurrency scheduling rule for controlling the calling sequence and concurrency number of the application layer interface.

[0115] The timeout fallback mechanism can be the processing logic when the interface call times out.

[0116] Specifically, the timestamp and request volume data in the historical traffic peak characteristics are analyzed to identify historical peak events, and the number of concurrent requests for each historical peak event is counted, and the maximum value of the maximum concurrent request number is set as the peak concurrency threshold.

[0117] Represent each application layer interface as a node (for example, node A represents the product query interface, and node B represents the inventory check interface). Based on the calling order in the interface call path characteristics, add directed edges between nodes (for example, the edge from node A to node B represents the product query interface calling the inventory check interface), thereby constructing an interface dependency graph.

[0118] Analyze the dependencies in the interface dependency graph and design concurrent calling rules. For interfaces without dependencies (such as independent nodes in the graph), set a parallel calling strategy (concurrent execution rules designed for nodes without dependencies in the interface dependency graph, such as allowing multiple independent interfaces to be called concurrently at the same time); for interfaces with dependencies (nodes connected by directed edges in the graph), set a serial calling strategy (sequential execution rules designed for nodes with dependencies in the interface dependency graph, such as calling the parent interface first and then the child interface); then, combined with the peak concurrency threshold, assign a concurrency upper limit to each interface to generate an interface call concurrency strategy.

[0119] Based on the interface dependency graph, identify potential timeout risk points (interface nodes in the interface dependency graph that are prone to response timeouts). Set a timeout threshold (maximum allowable response time) for each interface and generate a timeout fallback mechanism. For example, if an interface call times out, a fallback to a pre-set backup path is triggered. Integrate the interface call concurrency policy with the timeout fallback mechanism to determine the interface concurrency logic at the application layer.

[0120] Through this solution, the historical traffic peak characteristics are analyzed, the peak concurrency threshold is determined, the maximum concurrency pressure of the historical peak event is reflected, and the defect of being unable to predict the time attribute characteristics is resolved. The interface call path characteristics are analyzed, the interface dependency graph is determined, and the defect of being unable to parse the event attribute characteristics is eliminated. According to the peak concurrency threshold and the interface dependency graph, the interface call concurrency strategy and timeout fallback mechanism are generated to optimize the calling logic of the application layer interface, prevent congestion caused by multiple triggers of the same interface at the same time, and alleviate the pressure in high concurrency scenarios; at the same time, the response timeout or error problem is eliminated, and the entire link is avoided from being paralyzed due to the blocking of a single interface, reducing the deterioration of user experience. The interface call concurrency strategy and timeout fallback mechanism are used as the interface concurrency logic of the application layer to eliminate the defect of missing application layer interface concurrency logic.

[0121] In some embodiments, historical event data is parsed to obtain user behavior logs and network traffic logs; user behavior logs are analyzed to determine user behavior patterns; network traffic logs are analyzed to determine traffic fluctuation patterns; and event type classification rules are determined based on user behavior patterns and traffic fluctuation patterns.

[0122] User behavior logs can be structured log files parsed from historical event data, including user ID, operation type, timestamp, interface call path, etc.

[0123] Network traffic logs can be structured log files parsed from historical event data, including interface request volume, response time, data throughput, etc.

[0124] User behavior patterns can be high-frequency operation sequences or quantitative features of user preference distribution determined by analyzing user behavior logs.

[0125] The traffic fluctuation pattern may be a traffic peak feature or a quantitative feature of a fluctuation pattern determined by analyzing network traffic logs.

[0126] Specifically, parse the user operation records in historical event data (including user ID, operation type, timestamp, interface call path, etc.) to generate user behavior logs; parse the network traffic records in historical event data (including interface request volume, response time, data throughput, etc.) to generate network traffic logs.

[0127] Traverse the operation types and timestamps in the user behavior log to determine the high-frequency operation sequence (for example, count the number of occurrences of the product query → shopping cart addition sequence in the user behavior log); determine the user behavior pattern (such as user browsing preferences) based on the operation type. For example, by counting the proportion of operation types (such as browsing gift pages), high-frequency behaviors (user operations that occur frequently) can be identified; then, the high-frequency operation sequence and user behavior pattern are aggregated into a user behavior pattern.

[0128] Traverse the interface request volume and response time in the network traffic log, identify historical peak events, and locate sudden peaks (abnormal events in which the interface request volume suddenly increases significantly) by comparing the interface request volume with the baseline value (average request volume). Then, count the traffic fluctuation patterns (dynamic changes in network traffic characteristics) and mark periodic patterns (such as peaks occurring at fixed times of the day during holiday events). Then, summarize the sudden peaks and periodic patterns into traffic fluctuation patterns.

[0129] Match user behavior patterns with traffic fluctuation patterns (for example, if the user behavior pattern shows high-frequency concurrent interface calls and the traffic fluctuation pattern shows sudden peaks, they are associated with the same event). Based on the matching results, determine the event type classification rules. For example, if the user behavior pattern shows browsing preferences concentrated in a certain product category and the traffic fluctuation pattern shows periodic peaks, then classify it as a holiday event; if the user behavior pattern shows a high-concurrent interface call sequence and the traffic fluctuation pattern shows sudden peaks, then classify it as a flash sale event.

[0130] This solution analyzes historical event data to generate user behavior logs and network traffic logs. This allows for correlation between user behavior pattern analysis, supports quantification of traffic fluctuation patterns, and avoids coupling operation records with traffic metrics during analysis. By analyzing user behavior logs and identifying user behavior patterns, we address the inability to predict event attributes. By analyzing network traffic logs and identifying traffic fluctuation patterns, we address the inability to predict temporal attributes. Based on user behavior and traffic fluctuation patterns, we determine event classification rules, eliminating the core drawback of a lack of dynamic analysis capabilities.

[0131] Figure 3 A schematic diagram of a self-built network construction system provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the self-built network construction system 300 of this embodiment includes: a demand analysis module 301 , a characteristic determination module 302 , an event characteristic analysis module 303 , and a time characteristic analysis module 304 .

[0132] The demand analysis module 301 is used to obtain network construction requirements; analyze the network construction requirements, and determine the network usage scenario and network usage scale; A characteristic determination module 302 is configured to determine time attribute characteristics and event attribute characteristics based on the network usage scenario and the network usage scale; The event characteristics analysis module 303 is used to determine the front-end response priority and the interface concurrency logic of the application layer according to the event attribute characteristics; The time characteristic analysis module 304 is used to determine the response priority of the front end and the triggering timing of the interface concurrency logic of the application layer according to the time attribute characteristics.

[0133] Optionally, when the characteristic determination module 302 determines the time attribute characteristics and event attribute characteristics based on the network usage scenario and the network usage scale, it is used to: obtain historical event data, analyze the historical event data, and determine event type classification rules; based on the event type classification rules, analyze the network usage scenario to determine the network congestion event type; based on the network congestion event type, analyze the historical event data to determine the historical traffic peak characteristics and interface call path characteristics; determine the time attribute characteristics and event attribute characteristics based on the historical traffic peak characteristics and the interface call path characteristics.

[0134] Optionally, when the event characteristic analysis module 303 determines the response priority of the front end and the interface concurrency logic of the application layer according to the event attribute characteristics, it is used to: determine the user browsing core and browsing page content according to the event attribute characteristics; analyze the historical event data to determine the event browsing preference; split the browsing page content to obtain several content units; based on the event browsing preference, prioritize the several content units to obtain the response priority of the front end; generate the interface concurrency logic of the application layer according to the historical traffic peak characteristics and the interface call path characteristics.

[0135] Optionally, when the characteristic determination module 302 determines the time attribute characteristics and event attribute characteristics based on the historical traffic peak characteristics and the interface call path characteristics, it is used to: analyze the historical traffic peak characteristics to determine the periodic time pattern and the burst time pattern; analyze the interface call path characteristics to determine the interface response delay pattern and the interface error rate pattern; determine the time attribute characteristics and event attribute characteristics based on the periodic time pattern, the burst time pattern, the interface response delay pattern and the interface error rate pattern.

[0136] Optionally, the historical event data includes historical traffic data. The characteristic determination module 302 analyzes the network usage scenario based on the event type classification rule to determine the type of network congestion event, and is used to: extract the network usage scenario and determine the scenario characteristic parameters; analyze the historical traffic data and identify traffic anomalies; based on the event type classification rule, determine the initial matching type according to the traffic anomalies and the scenario characteristic parameters; verify the initial matching type in combination with the network usage scale to determine the type of network congestion event.

[0137] Optionally, when the event characteristic analysis module 303 analyzes the historical event data and determines the event browsing preference, it is used to: parse the historical event data to obtain user click stream data; analyze the user click stream data to determine the page dwell time distribution and the click heat map; determine the content attention index based on the page dwell time distribution and the click heat map; filter low-attention events in combination with the event attribute characteristics and the content attention index to determine the event browsing preference.

[0138] Optionally, the event characteristic analysis module 303 prioritizes several content units based on the event browsing preference, and when obtaining the front-end response priority, it is used to: assign a preference weight to each content unit according to the event browsing preference; analyze the historical traffic peak characteristics to determine the unit loading delay tolerance; calculate the priority score according to the preference weight and the unit loading delay tolerance; and sort the several content units based on the priority score to obtain the front-end response priority.

[0139] Optionally, when the event characteristic analysis module 303 generates the interface concurrency logic of the application layer based on the historical traffic peak characteristics and the interface call path characteristics, it is used to: analyze the historical traffic peak characteristics to determine the peak concurrency threshold; analyze the interface call path characteristics to determine the interface dependency graph; generate the interface call concurrency strategy and timeout fallback mechanism based on the peak concurrency threshold and the interface dependency graph; and use the interface call concurrency strategy and the timeout fallback mechanism as the interface concurrency logic of the application layer.

[0140] Optionally, when the characteristic determination module 302 analyzes the historical event data and determines the event type classification rules, it is used to: parse the historical event data to obtain user behavior logs and network traffic logs; analyze the user behavior logs to determine user behavior patterns; analyze the network traffic logs to determine traffic fluctuation patterns; and determine event type classification rules based on the user behavior patterns and the traffic fluctuation patterns.

[0141] The system of this embodiment can be used to execute the method of any of the above embodiments. Its implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A method for building a self-built network, characterized in that: include: Obtain network construction requirements; analyze the network construction requirements and determine the network usage scenarios and network usage scale; Determining time attribute characteristics and event attribute characteristics based on the network usage scenario and the network usage scale; Determine the front-end response priority and application layer interface concurrency logic based on the event attribute characteristics; According to the time attribute characteristics, the response priority of the front end and the triggering timing of the interface concurrency logic of the application layer are determined.

2. The method according to claim 1, characterized in that The determining of time attribute characteristics and event attribute characteristics based on the network usage scenario and the network usage scale includes: Acquire historical event data, analyze the historical event data, and determine event type classification rules; Analyzing the network usage scenario based on the event type classification rule to determine the type of network congestion event; Based on the type of network congestion event, analyze the historical event data to determine historical traffic peak characteristics and interface call path characteristics; According to the historical traffic peak characteristics and the interface call path characteristics, the time attribute characteristics and the event attribute characteristics are determined.

3. The method according to claim 2, characterized in that Determining the front-end response priority and the application layer interface concurrency logic based on the event attribute characteristics includes: Determine the user's browsing focus and browsing page content based on the event attribute characteristics; Analyze the historical event data to determine event browsing preferences; Splitting the browsing page content to obtain a plurality of content units; Based on the event browsing preference, a plurality of content units are prioritized to obtain a response priority of the front end; Generate application layer interface concurrency logic based on the historical traffic peak characteristics and the interface call path characteristics.

4. The method according to claim 2, characterized in that The determining of time attribute characteristics and event attribute characteristics based on the historical traffic peak characteristics and the interface call path characteristics includes: Analyze the historical traffic peak characteristics to determine periodic time patterns and sudden time patterns; Analyze the interface call path characteristics to determine the interface response delay pattern and interface error rate pattern; Time attribute characteristics and event attribute characteristics are determined according to the periodic time pattern, the burst time pattern, the interface response delay pattern, and the interface error rate pattern.

5. The method according to claim 2, characterized in that The historical event data includes historical traffic data, and the analyzing the network usage scenario based on the event type classification rule to determine the type of network congestion event includes: Extracting the network usage scenario and determining scenario characteristic parameters; Analyze the historical traffic data and identify traffic anomalies; Based on the event type classification rule, according to the traffic anomaly point and the scene feature parameters, determine the initial matching type; The initial matching type is verified in combination with the network usage scale to determine the type of network congestion event.

6. The method according to claim 3, characterized in that The analyzing the historical event data to determine event browsing preferences includes: Parsing the historical event data to obtain user click stream data; Analyze the user click stream data to determine page dwell time distribution and click heat map; Determining a content attention index based on the page dwell time distribution and the click heat map; Combined with the event attribute characteristics and the content attention index, low-attention events are filtered to determine event browsing preferences.

7. The method according to claim 3, characterized in that The prioritization of the plurality of content units based on the event browsing preference to obtain the front-end response priority includes: assigning a preference weight to each content unit based on the event browsing preference; Analyze the historical traffic peak characteristics and determine the unit loading delay tolerance; Calculating a priority score based on the preference weight and the unit loading delay tolerance; Based on the priority scores, the plurality of content units are sorted to obtain a response priority of the front end.

8. The method according to claim 3, characterized in that Generating the interface concurrency logic of the application layer according to the historical traffic peak characteristics and the interface call path characteristics includes: Analyze the historical traffic peak characteristics and determine the peak concurrency threshold; Analyze the interface call path characteristics and determine the interface dependency graph; Generate an interface call concurrency strategy and a timeout fallback mechanism based on the peak concurrency threshold and the interface dependency graph; The interface call concurrency strategy and the timeout fallback mechanism are used as the interface concurrency logic of the application layer.

9. The method according to claim 2, characterized in that The analyzing the historical event data to determine event type classification rules includes: Parsing the historical event data to obtain user behavior logs and network traffic logs; Analyze the user behavior log to determine user behavior patterns; Analyzing the network traffic logs to determine traffic fluctuation patterns; An event type classification rule is determined based on the user behavior pattern and the traffic fluctuation pattern.

10. A self-built network construction system, characterized in that: The method as claimed in any one of claims 1 to 9 comprises: A demand analysis module is used to obtain network construction requirements; analyze the network construction requirements, and determine the network usage scenario and network usage scale; a characteristic determination module, configured to determine time attribute characteristics and event attribute characteristics according to the network usage scenario and the network usage scale; An event characteristics analysis module is used to determine the front-end response priority and the interface concurrency logic of the application layer based on the event attribute characteristics; The time characteristic analysis module is used to determine the response priority of the front end and the triggering time of the interface concurrency logic of the application layer according to the time attribute characteristics.

Images