Method and system for processing network request data

Abstract

The application discloses a network request data processing method and processing system. The processing method comprises the following steps: capturing network request data; parsing the network request data into processing information; matching the processing information with processing modes in a preset rule library; if the matching is successful, automatically processing the network request according to the corresponding processing mode; if the matching fails, triggering manual processing of the network request. The technical scheme of the application can effectively improve the processing efficiency and meet the continuity requirement of the business.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a network request data processing method and system. BACKGROUND

[0002] Under the background of current digital office and popular network application, network request data processing has become the core link of various business systems.

[0003] The traditional method of processing network request data usually adopts manual operation. Although manual operation can flexibly cope with complex scenarios, it has low efficiency. Especially in high repetitive tasks, manual processing not only takes a long time, but also is prone to operation errors due to fatigue, inattention and other factors. Especially during peak business periods, the manual load increases sharply, and the processing delay and error rate further increase, which seriously affects business continuity. SUMMARY

[0004] In view of the defects in the prior art, the present application provides a network request data processing method which can effectively improve the processing efficiency and meet the continuity requirements of business.

[0005] The network request data processing method provided by the present application comprises:

[0006] grabbing network request data;

[0007] parsing the network request data into processing information;

[0008] matching the processing information with processing modes in a preset rule library;

[0009] if the matching is successful, automatically processing the network request according to the corresponding processing mode;

[0010] if the matching fails, triggering manual processing of the network request.

[0011] In one aspect, the step of triggering manual processing of the network request comprises:

[0012] saving the network request data to a manual task pool and sorting the network request data according to a preset strategy for processing in sequence.

[0013] In one aspect, the preset strategy includes sorting according to the urgency, importance, complexity, business category or timeliness of the network request data.

[0014] In one aspect, after the step of automatically processing the network request according to the corresponding processing mode, it comprises:

[0015] determining whether the automatic processing is successful;

[0016] If successful, submit the task;

[0017] If failed, transfer the network request data to the artificial task pool.

[0018] In one aspect, the step of triggering artificial processing of the network request comprises:

[0019] Obtaining an artificial processing strategy, converting the artificial processing strategy into a processing rule, and updating the processing rule to the preset rule library.

[0020] In one aspect, the step of matching the processing information with the processing mode in the preset rule library comprises:

[0021] Matching the processing information with the processing mode in the preset rule library to form a matching degree, and if the matching degree is less than a preset matching threshold, the matching is successful.

[0022] In one aspect, the step of automatically processing the network request according to the corresponding processing mode comprises:

[0023] Evaluating the effect of the automatically processed task, and if the processing accuracy of the corresponding processing mode is lower than a standard threshold, automatically triggering rule reconstruction.

[0024] In one aspect, the step of capturing the network request data comprises:

[0025] Judging whether the captured data is successful;

[0026] If the capturing fails, re-performing the capturing step;

[0027] If the capturing is successful, parsing the network request data.

[0028] In one aspect, the step of automatically processing the network request according to the corresponding processing mode further comprises:

[0029] Forming a feedback message, and based on the feedback message, continuing the step of capturing the network request data.

[0030] In addition, to solve the above problems, the application also provides a network request data processing system, comprising:

[0031] A capturing module for capturing network request data;

[0032] A parsing module for parsing the network request data into processing information;

[0033] A matching module for matching the processing information with a processing mode in a preset rule library;

[0034] an automatic processing module, configured to automatically process the network request according to the corresponding processing mode if the matching is successful;

[0035] a manual processing module, configured to trigger manual processing of the network request if the matching is unsuccessful.

[0036] The present application has the beneficial effects that: by capturing network request data and parsing into processing information, and matching the processing information with the processing modes in the preset rule library, the network request with successful matching is automatically processed, and the network request with unsuccessful matching is triggered for manual processing, effectively solving the problems of low efficiency, easy errors and high processing delay and error rate in high repetitive tasks in pure manual operation, improving the processing efficiency of high repetitive tasks by means of automatic processing, and coping with complex scenarios not covered by rules through manual processing, realizing the cooperation of automation and manual processing, thereby improving the overall processing efficiency and meeting the business continuity requirements. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the specific embodiments or prior art description will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual proportions.

[0038] Figure 1 a flow step schematic diagram of the network request data processing method of the present application;

[0039] Figure 2 a flow step schematic diagram of the network request data processing method of the present application after automatic processing fails;

[0040] Figure 3 a flow step schematic diagram of the network request data processing method of the present application for updating the preset rule library;

[0041] Figure 4 a flow step schematic diagram of the network request data processing method of the present application for automatically triggering rule reconstruction;

[0042] Figure 5 a flow step schematic diagram of the network request data processing method of the present application for judging whether the network request data is successfully captured;

[0043] Figure 6 a functional structure schematic diagram of the network request data processing system of the present application. DETAILED DESCRIPTION

[0044] The embodiments of the technical solutions of the present application will be described in detail below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0045] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in the present application should be understood as the usual meaning understood by the skilled person in the field to which the present application belongs.

[0046] As Figure 1 shown, the present application provides a network request data processing method, the processing method comprising:

[0047] Step S10, capturing network request data; through technical means to capture the network request data transmitted between the browser and the server, such as the order request submitted by the user on the e-commerce platform, the query request initiated in the financial system, etc., covering request parameters, request header, request method and other information. This operation replaces the manual data collection method, ensuring that the original data can be timely and completely included in the processing flow.

[0048] Among them, the capturing method can use browser extension API to intercept requests at the browser level; it can also forward traffic and record requests through a proxy server; or inject a hook script to rewrite the original request function to capture data; it can also use network layer packet capture tools to analyze underlying data packets to extract requests; collect received request information from the server log or use browser developer tool protocol to listen to and obtain requests.

[0049] Step S20, parsing the network request data into processing information; the original network request data captured is converted into structured information that meets the processing requirements. For example, the product ID, quantity, amount and other parameters in the order request are extracted from the original request message and arranged as order information: product A, quantity 2, total amount 500 yuan, etc. Clear processing information. The parsed processing information is not only convenient for machines to perform rule matching, but also convenient for manual quick understanding of task core, avoiding the processing obstacles caused by the complex format of original data, and improving the usability of data.

[0050] Step S30, matching the processing information with the processing mode in the preset rule library; by comparing the parsed processing information with the processing mode in the preset rule library, it is judged whether the current request can be automatically processed. For example, there is a processing mode in the preset rule library that when the order amount is less than or equal to 1000 yuan and the user is VIP, the delivery is automatically approved. If the processing information is that the order amount is 800 yuan and the user is VIP, it can be matched with the processing mode. This matching mechanism realizes the rapid classification of tasks, distinguishes between regular tasks that can be standardized and special tasks that need human intervention, provides a basis for hierarchical processing, and avoids the inefficiency of indiscriminate processing of all tasks in traditional manual processing.

[0051] Step S41, if the matching is successful, the network request is automatically processed according to the corresponding processing mode; for example, after matching the rule of automatically approving delivery, the system automatically sends a delivery instruction to the warehouse system and updates the order status to shipped. This process fully utilizes the advantages of machines in repetitive and standardized tasks, solves the problems of low efficiency and easy errors in pure manual processing of high repetitive tasks, greatly improves the processing speed and accuracy, and reduces the manual input in mechanical labor.

[0052] Step S42, if the matching fails, triggering manual processing of the network request. For example, if an order amount is 1500 yuan and the user is not VIP, there is no corresponding processing mode in the rule library, and the order information is displayed to the human auditor for judgment whether to approve delivery. This design retains the flexibility of human judgment for complex and special scenarios, makes up for the limited coverage of the preset rule library, avoids processing interruption caused by fixed rules, ensures the continuous progress of business, and solves the problem that single automated processing cannot balance efficiency and flexibility.

[0053] In this embodiment, by capturing network request data and parsing it into processing information, and after matching with the processing mode in the preset rule library, the network request that matches successfully is automatically processed, and the network request that fails to match triggers manual processing. This effectively solves the problems of low efficiency, easy errors, and high processing delay and error rate during peak business of pure manual operation in high repetitive tasks. It not only improves the processing efficiency of high repetitive tasks by means of automated processing, but also handles complex scenarios not covered by rules through manual processing, realizes the cooperation of automation and manual processing, thereby improving the overall processing efficiency and meeting the business continuity requirements.

[0054] In an embodiment of the present application, the step of triggering manual processing of the network request comprises:

[0055] Step S421: Save the network request data to the manual task pool, and sort the network request data according to a preset strategy for sequential processing. Store network request data that fails to match or fails to be automatically processed in the manual task pool to ensure that these tasks requiring manual intervention are centrally managed and not overlooked.

[0056] Furthermore, the preset strategy includes sorting network request data based on its urgency, importance, complexity, business category, or timeliness.

[0057] In terms of urgency, the network request data is sorted according to the urgency label or urgency status of the task. For example, order review requests marked with an emergency processing label and user complaint requests that require immediate response are given priority to human processing personnel to ensure that urgent tasks are responded to quickly.

[0058] In terms of importance, requests can be sorted according to the business value or scope of impact of the network request data. For example, requests involving core business have higher priority than ordinary query requests, ensuring the priority of processing important business.

[0059] For complexity, requests can be sorted according to the difficulty of processing network request data. For example, requests with complex data relationships, containing unstructured information, or fuzzy requests with low matching degree in the rule base can be sorted according to preset complexity levels, such as high, medium, and low, so that human processing personnel can reasonably allocate their efforts.

[0060] For business categories, requests can be grouped and sorted according to the business area to which the network request data belongs. For example, requests of different categories such as financial transactions, e-commerce orders, and system maintenance can be classified and sorted separately, so that human processors who are good at the corresponding areas can focus on handling similar tasks, thereby improving the professionalism and efficiency of processing.

[0061] Regarding timeliness, data requests can be sorted according to the processing time limits set for network requests. For example, user information update requests that require completion within 24 hours or report submission requests that need to be processed before the end of the workday can be sorted by the remaining time to avoid business delays due to timeouts.

[0062] like Figure 2 As shown, after the steps of automatically processing network requests according to the corresponding processing mode, the process includes:

[0063] Step S50 determines whether the automatic processing was successful. For example, based on preset success criteria, such as the processing result conforming to business rules and the response status being normal, it checks whether the automatic processing completed according to the matched processing mode achieved the expected results. This ensures the reliability of automated processing and avoids erroneous processing results taking effect directly due to rule loopholes or abnormal data, thus providing an error correction and gatekeeping mechanism.

[0064] Step S51: If successful, submit the task. When automatic processing is deemed successful, the completed task is formally submitted, allowing it to enter downstream business processes, such as submission to the warehousing system after order confirmation or updating the database after data verification. This operation marks the completion of the closed loop of the automated processing flow, fully leveraging the efficiency of machines in standardized tasks, directly achieving the goal of improving processing efficiency, and reducing manual intervention.

[0065] Step S52: If the process fails, the network request data is transferred to the manual task pool. If automatic processing fails, such as due to processing results not conforming to business rules or system execution errors, the corresponding network request data is transferred to the manual task pool and incorporated into the manual processing flow. Manual intervention compensates for the limitations of automated processing, avoids task stagnation caused by automatic processing failures, and ensures that even if problems occur in the automated process, tasks can still be resolved through manual processing, thus guaranteeing the continuity of the business process.

[0066] like Figure 3 As shown, after triggering the manual handling of a network request, the steps include:

[0067] Step S60: Obtain the manual processing strategy, convert the manual processing strategy into processing rules, and update the processing rules to the preset rule base. After manual processing is completed, automatically obtain the strategies formed during the manual processing, such as the judgment logic and operation steps of the manual for specific network requests; then, through a preset conversion mechanism, convert these manual strategies into machine-recognizable processing rules, such as structured conditions and action relationships; and then update the preset rule base with the newly generated processing rules. This allows the rule base to continuously absorb the experience of manual processing, gradually cover more complex scenarios, reduce the frequency of manual intervention for subsequent similar requests, and achieve a virtuous cycle of manual processing once, rule base update once, and subsequent automatic processing, continuously improving the system's automated processing capabilities and adaptability to business scenarios, and improving overall processing efficiency.

[0068] Furthermore, machine learning models can be used to transform manual processing strategies into processing rules. Specifically, after manual processing is completed, the system automatically collects the processing information of network requests, the manual processing strategy, and the processing results. These three elements form labeled training samples, with labels including processing conditions, operational actions, and business scenarios. Machine learning models, such as decision trees and association rule algorithms, are periodically invoked to perform cluster analysis on the training samples, extracting high-frequency processing logic. For example, when request parameters contain abnormal addresses and the amount exceeds 5000 yuan, manual review is required, generating candidate processing rules. The accuracy of candidate processing rules can be verified using historical data, and once the accuracy meets the standard, they are updated to a preset rule base. Thus, the machine learning model can automatically extract rules from manual processing cases.

[0069] In one embodiment of this application, the step of matching processing information with processing patterns in a preset rule base includes:

[0070] Step S310: Match the processing information with the processing patterns in the preset rule base to form a matching degree. If the matching degree is less than the preset matching threshold, the match is successful. The matching degree is calculated by comparing the parsed processing information with the characteristics of the processing patterns in the rule base, such as trigger conditions and business parameters. For example, when the processing information is an order amount of 800 yuan and a user level of VIP, compared with the processing patterns in the rule base where the order amount is less than or equal to 1000 yuan and the user level is VIP, the matching degree can be determined based on the degree of parameter matching. If the calculated matching degree is less than the preset matching threshold, meaning the difference between the processing information and the processing pattern is within an acceptable range, the match is considered successful, and the system enters the automatic processing stage. By setting a matching threshold, the applicable scope of automated processing can be flexibly controlled, ensuring the accuracy of rule matching while avoiding misjudgment of tasks that could be automatically processed as matching failures due to minor differences. This improves the system's coverage of standardized tasks and optimizes the efficiency of human-machine collaboration.

[0071] The preset matching threshold can be a dynamic threshold. When setting the preset matching threshold, you can set different matching thresholds based on the business scenario of the network request data, such as financial transactions, ordinary queries, form submissions, or data sensitivity, such as containing private information or public information. For example, the threshold can be set to 80% in the financial scenario and 60% in the ordinary query scenario.

[0072] Before matching and processing information and modes, the system automatically calls the corresponding matching threshold by parsing the business scenario tags of the current request. In other words, the preset matching threshold can be automatically and dynamically adjusted according to the usage scenario.

[0073] like Figure 4 As shown, after the steps of automatically processing network requests according to the corresponding processing mode, the process includes:

[0074] Step S70: Evaluate the effectiveness of the automatically processed task. If the processing accuracy of the corresponding processing mode is lower than the standard threshold, rule reconstruction is automatically triggered. After the automatic processing task is completed, the processing results are evaluated, such as determining whether the results meet actual business needs and whether there are any errors, and the processing accuracy of the corresponding processing mode is calculated. When the calculated accuracy is lower than the preset standard threshold, the rule reconstruction process is automatically triggered. This allows for the timely detection and correction of inefficient or error-prone processing modes, avoiding continuous errors caused by rigid rules, improving the reliability and accuracy of automatic processing through dynamic optimization of the rule base, and further enhancing the efficiency of human-machine collaborative processing.

[0075] For the rule reconstruction process, it can include the following steps: data review, root cause analysis, rule adjustment, and new rule verification.

[0076] Specifically, during data review, historical processing records corresponding to the processing mode are automatically extracted, including successful cases and error cases, as well as manual correction results of error cases, business scenario descriptions, and other information, forming a complete review data set.

[0077] During root cause analysis, through the review data set, the feature differences between successful and error cases are compared, and combined with business logic, the core reasons for rule failure are located, such as overly broad / narrow rule conditions, failure to cover newly emerging business variables, mismatch between matching logic and actual demand, etc., and the analysis results are output. At this time, you can also combine the manually annotated error causes to improve analysis accuracy.

[0078] During rule adjustment, the rule content is optimized according to the analysis results.

[0079] For example, if the condition setting is unreasonable, the trigger condition can be refined or relaxed, such as adding a sub-condition that the order amount is greater than 1000 yuan; if there are uncovered scenarios, new exception rules or expanded rule application scope can be added; if the matching logic is biased, the feature weight or matching algorithm can be adjusted, such as increasing the priority of user level in rule matching.

[0080] During new rule verification, the adjusted new rule is verified in the test environment, and its processing accuracy is tested with historical error cases and simulated new scenario data to ensure that it is not lower than the standard threshold, while avoiding causing new errors. After verification, replace the old rule in the rule library with the new rule, and record the rule change log; after going online, continuously monitor its actual processing effect to ensure that the reconstructed rule is stable and effective.

[0081] The rule reconstruction process realizes dynamic iteration of rules through problem positioning, targeted optimization, and verification closed loop, ensuring that it always adapts to business needs and maintains high accuracy of automatic processing.

[0082] As can be seen from the above, rule reconstruction is performed when the processing accuracy is lower than the standard threshold, i.e. after the problem occurs. In addition, rule reconstruction can also be performed proactively, such as during the effectiveness evaluation of automatic processing tasks, while monitoring external environmental variables, including changes in web DOM structure, business rule updates, and changes in user operation habits. Set multi-dimensional trigger thresholds for monitoring external environmental variables, such as: web DOM structure change rate greater than 10%, new business rules released, manual processing of new scenarios accounting for more than 20%. If any trigger threshold is met, rule reconstruction is triggered proactively, achieving proactive rule reconstruction.

[0083] In the present application, a network request can be automatically processed according to a rule base or manually processed, and the two are independently completed for processing the network request. In addition, for some network conditions, a semi-automatic processing flow can be performed, that is, the main work is processed according to the rule base, and a small amount of work is manually participated. A network request is processed by a machine and a human being. For example, after processing information and processing mode matching, if the matching degree is in a preset transition interval between a success threshold and a failure threshold, a semi-automatic processing flow is entered. Based on the processing mode with the highest matching degree, part of the standardized operation is automatically completed, such as pre-filling the form basic field, marking the key items that need to be manually confirmed, and generating a semi-processing result. The semi-processing result is pushed to the manual task pool, and the manual processing personnel only need to confirm or modify the key items to complete the final processing; after the processing is completed, the manual supplement operation is recorded and is used as a training sample to update the rule base. For another example, if the order amount matching degree is in the preset transition interval, the delivery address is automatically filled according to the rule base, and the human being only confirms the payment method.

[0084] As shown in Figure 5 After the step of capturing network request data, the method further includes:

[0085] In step S101, it is determined whether the captured data is successful. After capturing the network request data, whether the capturing operation achieves the target is checked by a preset success determination standard, such as whether the data is complete, whether the format meets the expectation, whether the core field is included, and the like. The reliability of the data processing chain is ensured by avoiding the waste of resources or processing errors caused by invalid data entering the analysis link.

[0086] In step S110, if the capturing fails, the capturing step is performed again. When it is determined that the capturing fails, a re-capturing mechanism is triggered, and the network request data capturing operation in step S10 is performed again. Thus, a retry mechanism is formed, which makes up for the accidental failure that may exist in single capturing, reduces the task stagnation caused by capturing failure, and ensures that the network request data can be effectively obtained.

[0087] In step S111, if the capturing is successful, the network request data is analyzed. Thus, it is ensured that the effectively captured data can be timely converted into processing information that can be used for rule matching, the processing flow is promoted in an orderly manner, the invalid data occupies the analysis resources is avoided, and the overall processing efficiency is improved, thereby laying a data foundation for subsequent automatic processing or manual processing.

[0088] In an embodiment of the present application, the step of automatically processing the network request according to the corresponding processing mode further includes:

[0089] Step S201, a feedback message is formed, and based on the feedback message, the step of capturing network request data is continued. After completing the automatic processing of the network request, a feedback message is generated according to the result of the automatic processing, which contains response information related to the current processing task, for example, the automatic processing result of the order request has confirmed that additional information is needed, etc. Then, based on this feedback message, the network request data capture operation of step S10 is continued to trigger, realizing the continuous capture of subsequent associated requests. It ensures that the capture and processing of network request data form a coherent dynamic flow, avoids the interruption of the flow after single task processing, and further guarantees the continuity of business processing.

[0090] As shown in Figure 6 The application also provides a network request data processing system, which comprises a capture module, an analysis module, a matching module, an automatic processing module and a manual processing module.

[0091] The capture module is used for capturing network request data; the capture module is the data source entrance of the entire processing system, and captures the network request related information transmitted between the browser and the server through technical means, to provide original data support for subsequent processing. It ensures that the initial materials required for data processing are obtained.

[0092] The analysis module is used for analyzing network request data into processing information; the function of the analysis module is to analyze the network request data obtained by the capture module into processing information. The original network request data may contain complex formats or redundant contents, and the analysis module converts these data into structured information convenient for system recognition and processing through processing logic, so that it can meet the requirements of matching with the preset rule library, and provide processing information that can be directly used for subsequent matching.

[0093] The matching module is used for matching the processing information with the processing mode in the preset rule library; the role of the matching module is to match the processing information output by the analysis module with the processing mode in the preset rule library. By comparing the processing information with the processing mode defined in the rule library, it is judged whether they are matched, so as to decide whether the network request should enter the automatic processing flow or the manual processing flow.

[0094] The automatic processing module is used for automatically processing the network request according to the corresponding processing mode if the matching is successful; the automatic processing module starts to work when the matching module determines that the processing information and the processing mode in the preset rule library are matched successfully, and mainly automatically processes the network request according to the corresponding processing mode. The automatic processing module executes specific processing actions according to the preset operation logic in the rule library, and can complete the standardized and repetitive network request processing task without human intervention, effectively improving the processing efficiency and exerting the advantage of machine in repetitive labor.

[0095] The artificial processing module is used to trigger the artificial processing of the network request if the matching fails. The artificial processing module is triggered to run when the matching module determines that the processing information does not match the processing mode in the preset rule library, and starts the artificial processing flow of the network request. For complex or special scenarios not covered by the rule library, the artificial processing module directs the network request data to the artificial processing channel, and completes the processing by means of artificial judgment and operation, making up for the lack of flexibility of fixed rules, ensuring that all network requests can be effectively processed, and guaranteeing business continuity.

[0096] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the specification of the present application.

Claims

1. A method for processing network request data, characterized in that, The processing method includes: Capture network request data; The network request data is parsed into processing information; The processing information is matched with the processing patterns in the preset rule base; If a match is found, the network request will be processed automatically according to the corresponding processing mode. If a match fails, manual processing of the network request is triggered. After triggering the manual handling of the network request, the steps include: The process involves: acquiring manual processing strategies, converting these strategies into processing rules, and updating these rules to a preset rule base; including: automatically collecting processing information of network requests, manual processing strategies, and processing results; forming labeled training samples from these three elements; periodically calling a machine learning model to perform cluster analysis on the training samples; generating candidate processing rules based on the results of the cluster analysis (which are then manually reviewed); verifying the accuracy of the candidate processing rules using historical data; and updating these rules to the preset rule base once the accuracy meets the required standard. The step of matching the processing information with processing patterns in a preset rule base includes: The processing information is matched with the processing patterns in the preset rule base to form a matching degree. If the matching degree is less than the preset matching threshold, the match is successful. The matching threshold is dynamically adjusted according to the business scenario of the network request data. After automatically processing network requests according to the corresponding processing mode, the steps include: The effectiveness of the automatically processed tasks is evaluated. If the processing accuracy of the corresponding processing mode is lower than the standard threshold, rule reconstruction is automatically triggered. The rule reconstruction includes at least data review, root cause analysis, rule adjustment and new rule verification, so as to dynamically iterate the rules in the preset rule base.

2. The processing method according to claim 1, characterized in that, The steps that trigger manual processing of a network request include: The network request data is saved to a manual task pool and sorted according to a preset strategy for sequential processing.

3. The processing method according to claim 2, characterized in that, The preset strategy includes sorting network request data according to its urgency, importance, complexity, business category, or timeliness.

4. The processing method according to claim 2, characterized in that, After automatically processing network requests according to the corresponding processing mode, the steps include: Determine whether the automatic processing was successful; If successful, submit the task; If the request fails, the network request data will be transferred to the artificial task pool.

5. The processing method according to claim 1, characterized in that, After the steps of fetching network request data, the following are included: Determine whether the data retrieval was successful; If the crawling fails, repeat the crawling steps. If the capture is successful, the network request data will be parsed.

6. The processing method according to claim 1, characterized in that, The steps for automatically processing network requests according to the corresponding processing mode also include: A feedback message is generated, and the process of capturing network request data continues based on the feedback message.

7. A system for processing network request data, characterized in that, The processing system includes: The crawling module is used to crawl network request data; The parsing module is used to parse the network request data into processing information; The matching module is used to match the processing information with the processing patterns in the preset rule base; The automatic processing module is used to automatically process network requests according to the corresponding processing mode if a match is found. The manual processing module is used to trigger manual processing of network requests if a match fails. After triggering the manual handling of the network request, the steps include: The process involves: acquiring manual processing strategies, converting these strategies into processing rules, and updating these rules to a preset rule base; including: automatically collecting processing information of network requests, manual processing strategies, and processing results; forming labeled training samples from these three elements; periodically calling a machine learning model to perform cluster analysis on the training samples; generating candidate processing rules based on the results of the cluster analysis (which are then manually reviewed); verifying the accuracy of the candidate processing rules using historical data; and updating these rules to the preset rule base once the accuracy meets the required standard. The step of matching the processing information with processing patterns in a preset rule base includes: The processing information is matched with the processing patterns in the preset rule base to form a matching degree. If the matching degree is less than the preset matching threshold, the match is successful. The matching threshold is dynamically adjusted according to the business scenario of the network request data. After automatically processing network requests according to the corresponding processing mode, the steps include: The effectiveness of the automatically processed tasks is evaluated. If the processing accuracy of the corresponding processing mode is lower than the standard threshold, rule reconstruction is automatically triggered. The rule reconstruction includes at least data review, root cause analysis, rule adjustment and new rule verification, so as to dynamically iterate the rules in the preset rule base.

Images