An intelligent recognition system for information technology consulting

Abstract

The present invention discloses an intelligent recognition system for information technology consulting, which relates to the technical field of interface call performance management. The system automatically collects and analyzes the data input / output process and interface call information through a data collection module, a data analysis module, an interactive performance evaluation module, and an intelligent optimization module, extracts key features through machine learning algorithms, and calculates performance indicators such as the data collection frequency, signal jitter, and error retry rate of the interface. Based on this data, the system can evaluate the interactive performance of the interface in real time, compare it with the preset performance threshold, and automatically identify the groups of qualified and abnormal interfaces. It adjusts the data collection frequency, optimizes the interface call mechanism, adjusts the data flow path, etc. accordingly to improve the response speed and resource utilization rate of the system. During peak periods, the system can effectively control the interface call stacking rate and edge data processing rate to ensure the interactive efficiency in the information technology consulting scenario.

Description

Technical Field

[0001] The present invention relates to the technical field of interface call performance management, and particularly to an intelligent recognition system for information technology consulting. Background Art

[0002] In the field of modern information technology consulting, as a data exchange bridge between system modules, interfaces carry a large number of tasks and data transmissions. With the increasing complexity of business requirements, the performance of system interfaces has become a key factor affecting the overall system efficiency and user experience. Traditional information technology consulting systems often adopt static or simple optimization methods to manage interface performance, but these methods tend to be inadequate when faced with the growing system load and complex business scenarios.

[0003] In the environment of information technology consulting, the efficiency of interface calls directly affects the system response speed and resource utilization. When the call stacking rate of an interface is relatively high, the system will face problems such as increased latency, decreased processing capacity, and may even lead to interface timeouts, data loss, or system crashes. These problems generally stem from two aspects: one is the complexity and large volume of the data flow, and the other is the frequency of interface calls and error retries.

[0004] With the diversification of user requirements and the improvement of real-time requirements, traditional technologies usually rely on manual intervention or simple preset rules to handle interface performance problems, such as setting the tolerance of an interface through fixed thresholds, or relying on manual adjustment of data collection frequency and error retry mechanisms. The limitation of this approach is that it cannot dynamically adjust parameters according to the real-time system state and changes, nor does it have the ability of intelligent evaluation and optimization. In the face of a complex system environment, static rules often lead to over-optimization or resource waste, affecting the flexibility and reliability of the system. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the present invention provides an intelligent recognition system for information technology consulting to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent recognition system for information technology consulting includes a data collection module, a data analysis module, an interactive performance evaluation module, and an intelligent optimization module;

[0007] The data collection module is used to pre-define the data flow range in the information technology consulting scenario, and collect relevant data input-output processes and interface call information to establish an interactive dependency relationship data set;

[0008] The data analysis module is used to extract features from the input-output process, the performance characteristics of the i-th interface call, and the data dependency relationship through machine learning algorithms based on the interactive dependency relationship dataset, establish a data flow model and train it to calculate and obtain: the dynamic adjustment coefficient dttz of the data collection frequency of the i-th interface i and the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i ;

[0009] The interactive performance evaluation module is used to collect the number of hops of the data flow path transmitted by the p-th path of the i-th interface and the latency of the p-th path of the i-th interface , and associate the dynamic adjustment coefficient dttz of the data collection frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i to construct an interactive performance comprehensive index Zh i ; At the same time, a performance threshold Cp is preset to generate a comparison grading result between the interactive performance comprehensive index Zh i and the performance threshold Cp;

[0010] The intelligent optimization module is used to dynamically adjust the data collection frequency, optimize the interface call mechanism, or reallocate the data flow path according to the comparison grading result, including adjusting the interface call stacking rate during peak periods and optimizing the edge data processing rate, so as to improve the interactive efficiency in the information technology consulting scenario.

[0011] Preferably, the data collection module includes an input data collection unit, an output data collection unit, and a call relationship collection unit;

[0012] The input data collection unit is used to collect user interaction behaviors, record operation paths and input data, and intercept and parse data packets returned by third-party services;

[0013] The output data collection unit is used to collect the content transmitted by external interfaces and data status;

[0014] The call relationship collection unit is used to collect relevant data input-output processes and interface call information. The relevant data input-output processes include: the flow path and intermediate node information of data across modules, the loss rate during data flow, the number of changes during data flow, and the number of verification times during data flow;

[0015] The interface call information includes: the interface path record of module A calling module B, interface name, call order, dependency relationship chain, call frequency, and call data traffic;

[0016] Summarize the data collected by the input data acquisition unit, output data acquisition unit, and call relationship acquisition unit, and establish an interactive dependency relationship data set.

[0017] Preferably, the data analysis module includes a preprocessing unit, a model training unit, and a feature extraction unit;

[0018] The preprocessing unit is used to remove irrelevant data, outliers, duplicate data, and normalize the interactive dependency relationship data set to make it meet the requirements of the machine learning model. Then, the model training unit uses machine learning algorithms to train the interactive dependency relationship data set to establish a data flow model, and the feature extraction unit extracts the input signal features, output data features, data flow features, and call performance features of the i-th interface to analyze and calculate to obtain: the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i ;

[0019] The call performance features include the acquisition frequency, CPU occupancy ratio, and number of errors.

[0020] Preferably, the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i is calculated and obtained through the following formula:

[0021]

[0022] where represents the frequency change value of the acquisition frequency of the i-th interface in the current time period minus the acquisition frequency of the previous time period, represents the average acquisition frequency of the i-th interface, represents the CPU occupancy ratio of the i-th interface, represents the average CPU occupancy ratio of the i-th interface, represents the number of errors of the i-th interface, represents the maximum tolerable number of errors of the i-th interface, obtained according to the fault tolerance standard predetermined by the service level agreement; , and represent weight values;

[0023] The acquisition method of the input signal jitter value ddz of the i-th interface i is specifically as follows:

[0024] S11. Collect the input signal time series data of the i-th interface, and the expression is: ; n represents the total number of points of the input signal features;

[0025] S12. Calculate the standard deviation of the input signal characteristics of the i-th interface to obtain the signal change amplitude value :

[0026]

[0027]

[0028] Among them, represents the input signal characteristics of the i-th interface at the j-th moment, represents the mean value of the input signal characteristics of the i-th interface;

[0029] S13. According to the mean value of the input signal characteristics of the i-th interface and the signal change amplitude value i :

[0030]

[0031] The error retry rate cwl of the i-th interface i is calculated by the following formula:

[0032]

[0033] Among them, m represents the number of error categories, including network errors, interface timeouts, and server response failure categories;

[0034] represents the weight of each error type, represents the number of retries of the i-th interface for the k-th error type, represents the total number of failures of the i-th interface for the k-th error type.

[0035] Preferably, the interaction performance evaluation module includes a path hop count calculation unit and a delay calculation unit for each path;

[0036] The path hop count calculation unit is used to collect the number of intermediate nodes in the transmission process of the p-th path of the i-th interface, and calculate the data stream path hop count of the p-th path transmission of the i-th interface through the following formula :

[0037]

[0038] Among them, is the total number of intermediate nodes in the transmission process of the p-th path transmission, indicates that node g is a transit node on path p, otherwise , if the p-th path is a direct connection, i.e., there is no intermediate node, then ;

[0039] Each path delay calculation unit is used to calculate the delay of the p-th path of the i-th interface :

[0040]

[0041] Wherein, is the queue delay of the g-th node on the path, is the transmission delay of the link between node g and node g + 1.

[0042] Preferably, the interaction performance evaluation module further includes an association unit and an evaluation unit;

[0043] The association unit is used to calculate the data stream path hop count transmitted by the p-th path of the i-th interface , the delay of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i , after dimensionless processing, the comprehensive interaction performance index Zh of the i-th interface is calculated and obtained through the following formula i :

[0044]

[0045] Wherein, , , , and respectively represent the data stream path hop count transmitted by the p-th path of the i-th interface , the delay of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i 's weight value, and .

[0046] Preferably, the evaluation unit is used to preset a performance threshold Cp, and is used to generate a comparison grading result between the comprehensive interaction performance index Zh of the i-th interface i and the performance threshold Cp, including:

[0047] When the comprehensive interaction performance index Zh of the i-th interface iWhen it is greater than the performance threshold Cp, it indicates that the interaction performance of the i-th interface is qualified, and a first port qualified group is generated.

[0048] When the comprehensive interaction performance index Zh of the i-th interface satisfies the performance threshold Cp × 80% ≤ Zh i ≤ Cp, it indicates that the interaction performance of the i-th interface is unqualified, and a second port exception group is generated.

[0049] When the comprehensive interaction performance index Zh of the i-th interface i <Cp × 80%, it indicates that the interaction performance of the i-th interface is unqualified, and a third port exception group is generated.

[0050] Preferably, the intelligent optimization module includes a first optimization unit, a second optimization unit, and a third optimization unit.

[0051] After the first optimization unit identifies the first port qualified group, it triggers the first strategy, including: for the i-th port, increasing the current data collection frequency by 5% - 10%, improving the interface call stacking rate control mechanism by 10% - 15%; increasing the current edge data processing rate by 5% - 10%; maintaining the existing path without adjusting the data flow path hop count.

[0052] Preferably, after the second optimization unit identifies the second port exception group, it triggers the second strategy, including: for the i-th port, reducing the current data collection frequency by 5% - 10%, reducing the interface call stacking rate control mechanism by 10% - 15%; increasing the edge data processing rate by 5% - 10%; adjusting and reducing the data flow path hop count by 10% - 15%.

[0053] Preferably, after the third optimization unit identifies the third port exception group, it triggers the third strategy, including: for the i-th port, reducing the current data collection frequency by 20% - 30%, reducing the interface call stacking rate control mechanism by 16% - 30%; increasing the edge data processing rate by 11% - 20%; adjusting and reducing the data flow path hop count by 16% - 30%, and optimizing the error retry rate control mechanism by 20% - 30%.

[0054] The present invention provides an intelligent recognition system for information technology consulting, which has the following beneficial effects:

[0055] (1) For the intelligent recognition system for information technology consulting, through the predefined information technology consulting scenario, the data collection module can effectively identify and collect relevant input and output processes and interface call information, thus providing a high-quality data source for subsequent data analysis. Establishing an interaction dependency relationship data set provides key data support for the subsequent analysis of the system, helps to reveal the data flow and dependency relationships between interfaces, and provides a basis for formulating optimization strategies.

[0056] (2) The intelligent recognition system for information technology consulting can, through machine learning algorithms, enable the data analysis module to automatically extract key features from a large amount of data, helping to accurately establish a data flow model, thereby better capturing potential performance problems of the system. Calculate the data acquisition frequency dynamic adjustment coefficient dttz of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i provide a quantitative basis for subsequent performance evaluation and optimization, ensuring that the optimization strategy can be adjusted according to the actual system state.

[0057] (3) The intelligent recognition system for information technology consulting can, by calculating the number of data flow path hops transmitted by the p-th path of the i-th interface , accurately calculating the number of path hops helps to identify high-latency paths and provides a direction for subsequent optimization. The delay calculation not only considers the queue delay of each node but also includes the link transmission delay between nodes. The queue delay reflects the waiting time when data is processed at each node, while the link delay is related to factors such as the bandwidth and transmission distance of the network connection. The two together determine the total delay of the path.

[0058] (4) The intelligent recognition system for information technology consulting can, by integrating multi-dimensional performance indicators, enable the interactive performance comprehensive index to accurately reflect the actual performance of the interface, providing a reliable basis for subsequent optimization decisions. The function of the evaluation unit is to compare the interactive performance comprehensive index with the performance threshold Cp based on a preset performance threshold Cp, thereby generating different port groups. In this process, according to the comparison and grading results of the performance comprehensive index and the performance threshold Cp, the interfaces are divided into three different performance groups, and corresponding strategies are generated for dynamic optimization. Through the adjustment of different strategies, the system parameters can be refined under different performance conditions, avoiding resource waste or performance degradation caused by over-adjustment. Whether it is optimizing edge data processing, reducing the number of path hops, or adjusting the data acquisition frequency, the optimization strategy aims to improve the stability, response speed, and processing ability of the system. Description of the Drawings

[0059] Figure 1 is a schematic flow chart of an intelligent recognition system for information technology consulting according to the present invention. Detailed Embodiments

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0061] Embodiment 1

[0062] Please refer to Figure 1 , the present invention provides an intelligent recognition system for information technology consulting, including a data acquisition module, a data analysis module, an interaction performance evaluation module, and an intelligent optimization module;

[0063] The data acquisition module is used to pre-define the data flow range in the information technology consulting scenario, and collect relevant data input and output processes and interface call information to establish an interaction dependency relationship data set;

[0064] The data analysis module is used to extract features of the input and output processes, the performance characteristics of the i-th interface call, and the data dependency relationship based on the interaction dependency relationship data set through machine learning algorithms, establish a data flow model and train it to calculate and obtain: the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i ;

[0065] The interaction performance evaluation module is used to collect the number of data flow path hops transmitted by the p-th path of the i-th interface and the delay of the p-th path of the i-th interface , and associate the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i to construct an interaction performance comprehensive index Zh i ; At the same time, a performance threshold Cp is preset to generate a comparison grading result of the interaction performance comprehensive index Zh i and the performance threshold Cp;

[0066] The intelligent optimization module is used to dynamically adjust the data acquisition frequency, optimize the interface call mechanism or re-allocate the data flow path according to the comparison grading result, including adjusting the interface call stacking rate during peak periods and optimizing the edge data processing rate, so as to improve the interaction efficiency in the information technology consulting scenario.

[0067]

[0068] In this embodiment, through the data acquisition module and the data analysis module, the system can automatically collect and analyze various performance data of interface calls, such as data acquisition frequency, input signal jitter, error retry rate, etc. Using machine learning algorithms to extract features from these data, combined with the interaction dependency relationship, a data flow model is established, and the comprehensive interaction performance index Zh is calculated in real time i to make the performance evaluation more accurate and intelligent. Traditional methods usually rely on manual intervention or static rules to adjust interface performance, which not only easily leads to over-optimization or resource waste, but also is difficult to cope with complex real-time changes. By adopting the intelligent optimization module, based on the comparison between the real-time evaluated Zhi value and the performance threshold Cp, the system can automatically identify the performance bottleneck of the interface, and dynamically adjust the data acquisition frequency, optimize the interface call mechanism or reallocate the data flow path according to the actual situation. This dynamic adjustment method can effectively improve the system's response speed, reduce latency, and avoid problems such as timeouts and system crashes. The system can monitor in real time and adjust the call stacking rate and edge data processing rate of the interface according to the performance changes of the interface to adapt to the changes in business requirements and system load. Through intelligent optimization, the system can not only improve the current interaction efficiency, but also better adapt to the diversity and variability of business scenarios, ensuring the flexibility and reliability of the system. In traditional methods, interface performance optimization requires a large amount of manual adjustment and monitoring, while the intelligent optimization system can greatly reduce the need for manual intervention. The system can automatically make adjustments based on real-time data, helping the technical team better focus on higher-level strategic decisions and business innovation, and improving the overall work efficiency. This system can not only optimize in the initial stage, but also has the ability of continuous learning and adaptive adjustment. With the changes in business requirements and the progress of technology, the system can continuously adjust the optimization strategy to maintain its optimal performance in a dynamic environment. This self-adaptability ensures the stability and high efficiency of the system during long-term operation

[0069] Embodiment 2

[0070] This embodiment is an explanatory description based on Embodiment 1. Specifically, please refer to Figure 1 wherein the data acquisition module includes an input data acquisition unit, an output data acquisition unit, and a call relationship acquisition unit

[0071] The input data acquisition unit is used to collect user interaction behaviors, record operation paths and input data, and intercept and parse data packets returned by third-party services

[0072] The output data acquisition unit is used to collect external interface transmission content and data status

[0073] The call relationship collection unit is used to collect relevant data input and output processes and interface call information. The relevant data input and output processes include: the flow path and intermediate node information of data across modules, the loss rate during data flow, the number of changes during data flow, and the number of verification times during data flow;

[0074] The interface call information includes: the interface path record of module A calling module B, interface name, call order, dependency relationship chain, call frequency, and call data flow;

[0075] Summarize the data collected by the input data collection unit, output data collection unit, and call relationship collection unit to establish an interactive dependency relationship data set.

[0076] In this embodiment, the input data collection unit can not only record the user's interaction behavior, but also intercept and parse the data packets returned by third-party services, so as to obtain more comprehensive system data and provide rich background information for subsequent analysis.

[0077] The output data collection unit is used to collect the transmission content and data status of external interfaces to ensure that the interaction between the system and external interfaces is completely captured and provide a deep understanding of the communication process between the system and the outside world.

[0078] The collection of interface call information can detail the dependency relationships, call paths, interface names, call orders, call frequencies, and call data flows between modules. These information are crucial for subsequent interface performance evaluation, helping the system comprehensively understand the interaction relationships between modules and the usage of interfaces, so as to identify interfaces with high load, frequent calls, or excessive resource occupation. Based on the interactive dependency relationship data set, the system can accurately evaluate the performance of interfaces and data flow paths, identify problems such as high latency and high packet loss during data transmission, and provide specific directions for subsequent optimization. For example, some data flow paths may be transmitted between multiple modules and involve a high packet loss rate or data changes. The system can adjust the path and optimize the transmission mechanism to reduce problems.

[0079] Embodiment 3

[0080] This embodiment is an explanatory description based on Embodiment 1. Specifically, please refer to Figure 1 and the data analysis module includes a preprocessing unit, a model training unit, and a feature extraction unit;

[0081] The preprocessing unit is used to remove irrelevant data, outliers, duplicate data, and normalize the interactive dependency relationship dataset to make it meet the requirements of the machine learning model. Then, the model training unit uses machine learning algorithms to train the interactive dependency relationship dataset to establish a data flow model. The feature extraction unit extracts the input signal features, output data features, data flow features, and call performance features of the i-th interface to analyze and calculate to obtain: the dynamic adjustment coefficient dttzi of the data acquisition frequency of the i-th interface, the input signal jitter value ddzi of the i-th interface, and the error retry rate cwli of the i-th interface.

[0082] The call performance features include the acquisition frequency, CPU occupancy ratio, and the number of errors.

[0083] The dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i is obtained by calculating through the following formula:

[0084]

[0085] where represents the frequency change value obtained by subtracting the acquisition frequency of the i-th interface in the previous time period from the acquisition frequency in the current time period, represents the average acquisition frequency of the i-th interface, represents the CPU occupancy ratio of the i-th interface, represents the average CPU occupancy ratio of the i-th interface, represents the number of errors of the i-th interface, represents the maximum tolerable number of errors of the i-th interface, which is obtained according to the fault tolerance standard predetermined by the service level agreement; 、 and represent weight values; based on factors such as the acquisition frequency change, CPU occupancy ratio, and the number of errors of the interface, the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface is calculated i . This coefficient can reflect the load situation of the interface and adjust the data acquisition frequency according to actual needs, avoiding resource waste caused by excessive acquisition or response lag caused by too low acquisition frequency. It can flexibly adjust the resource allocation and performance of the interface without adding too much additional overhead.

[0086] The acquisition method of the input signal jitter value ddz of the i-th interface i is specifically as follows:

[0087] S11. Collect the input signal time series data of the i-th interface, and the expression is: ; n represents the total number of points of the input signal features;

[0088] S12. Calculate the standard deviation of the input signal characteristics of the i-th interface to obtain the signal change amplitude value :

[0089]

[0090]

[0091] where represents the input signal characteristics of the i-th interface at the j-th moment, represents the mean value of the input signal characteristics of the i-th interface;

[0092] S13. Based on the mean value of the input signal characteristics of the i-th interface and the signal change amplitude value i :

[0093]

[0094] This calculation method helps to identify the instability of the input signal and can provide detailed performance analysis for subsequent interface optimization, especially when signal fluctuations affect the interface performance.

[0095] The error retry rate cwl of the i-th interface i is calculated by the following formula:

[0096]

[0097] where m represents the number of error categories, including network errors, interface timeouts, and server response failure categories;

[0098] represents the weight of each error type, represents the number of retries of the i-th interface for the k-th error type, represents the total number of failures of the i-th interface for the k-th error type. Considering that different types of errors may affect the interface performance, the system calculates the retry rate of each type of error by setting different weights to ensure that the impact of different error types on the overall system performance is comprehensively evaluated. Through this calculation method, the system can identify interfaces with high retry rates and optimize these interfaces key to improve the stability and response speed of the overall system.

[0099] In this embodiment, by calculating in real time and dynamically adjusting the data acquisition frequency, signal jitter value, and error retry rate of the interface, the system can adjust its performance parameters according to the real-time interface load and status. This adaptive ability enables the system to cope with changing business requirements and dynamic system loads. Based on the dynamically adjusted coefficients and parameters, the system can automatically optimize the interface call mechanism, data acquisition frequency, and error retry strategy, thereby improving the overall efficiency of the system, reducing latency, minimizing resource waste, and enhancing the user experience.

[0100] Embodiment 4

[0101] This embodiment is an explanatory illustration based on Embodiment 1. Specifically, please refer to Figure 1 , the interactive performance evaluation module includes a path hop count calculation unit and a delay calculation unit for each path;

[0102] The path hop count calculation unit is used to collect the number of intermediate nodes in the transmission process of the p-th path of the i-th interface, and calculate the data stream path hop count of the p-th path of the i-th interface through the following formula :

[0103]

[0104] Among them, is the total number of intermediate nodes in the transmission process of the p-th path, indicates that node g is a transfer node on path p, otherwise , if the p-th path is a direct connection, that is, there is no intermediate node, then ;

[0105] The delay calculation unit for each path is used to calculate the delay of the p-th path of the i-th interface :

[0106]

[0107] Among them, is the queue delay of the g-th node on the path, is the transmission delay of the link between node g and node g + 1.

[0108] In this embodiment, by calculating the hop count of the path, the complexity of the path and the potential latency during data transmission can be reflected. A path with a higher hop count may imply more intermediate nodes, which may increase the transmission time and reduce the response efficiency. Therefore, accurately calculating the path hop count helps identify high-latency paths and provides a direction for subsequent optimization. The latency calculation not only considers the queue latency of each node but also includes the link transmission latency between nodes. The queue latency reflects the waiting time when data is processed at each node, while the link latency is related to factors such as the bandwidth and transmission distance of the network connection. The two together determine the total latency of the path.

[0109] Embodiment 5

[0110] This embodiment is an explanatory description based on Embodiment 4. Specifically, please refer to Figure 1 , the interactive performance evaluation module further includes an association unit and an evaluation unit;

[0111] The association unit is used to make the data stream path hop count of the p-th path transmitted by the i-th interface , the latency of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i , after dimensionless processing, the comprehensive interactive performance index Zh of the i-th interface is calculated through the following formula i :

[0112]

[0113] Among them, , , , and respectively represent the data stream path hop count of the p-th path transmitted by the i-th interface , the latency of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i weight values, and .

[0114] The evaluation unit is used to preset a performance threshold Cp, and is used to generate a comparison grading result between the comprehensive interactive performance index Zh of the i-th interface i and the performance threshold Cp, including:

[0115] When the comprehensive interaction performance index Zh of the i-th interface i > the performance threshold Cp, it indicates that the interaction performance of the i-th interface is qualified, and a first port qualified group is generated;

[0116] When the performance threshold Cp × 80% ≤ the comprehensive interaction performance index Zh of the i-th interface i ≤ the performance threshold Cp, it indicates that the interaction performance of the i-th interface is unqualified, and a second port abnormal group is generated;

[0117] When the comprehensive interaction performance index Zh of the i-th interface i < the performance threshold Cp × 80%, it indicates that the interaction performance of the i-th interface is unqualified, and a third port abnormal group is generated.

[0118] In this embodiment, by integrating multi-dimensional performance indicators, the comprehensive interaction performance index can accurately reflect the actual performance of the interface, providing a reliable basis for subsequent optimization decisions. The function of the evaluation unit is to compare the comprehensive interaction performance index with the performance threshold CP based on the preset performance threshold Cp, thereby generating different port groups. This process classifies the interfaces into three different performance groups according to the comparison results between the comprehensive performance index and the threshold.

[0119] Embodiment 6

[0120] This embodiment is an explanatory description carried out in Embodiment 5. Specifically, please refer to Figure 1 , the intelligent optimization module includes a first optimization unit, a second optimization unit, and a third optimization unit;

[0121] The first optimization unit is used to trigger the first strategy after identifying the first port qualified group, including: for the i-th port, increasing the current data collection frequency by 5% - 10%, improving the interface call stacking rate control mechanism by 10% - 15%; increasing the current edge data processing rate by 5% - 10%; maintaining the existing path without adjusting the data flow path hop count.

[0122] The second optimization unit is used to trigger the second strategy after identifying the second port abnormal group, including: for the i-th port, reducing the current data collection frequency by 5% - 10%, reducing the interface call stacking rate control mechanism by 10% - 15%; increasing the edge data processing rate by 5% - 10%; adjusting and reducing the data flow path hop count by 10% - 15%.

[0123] After identifying the third - port exception group, the third optimization unit triggers the third strategy, including: for the i - th port, reducing the current data - acquisition frequency by 20% - 30%, reducing the interface - call stacking - rate control mechanism by 16% - 30%; increasing the edge - data processing rate by 11% - 20%; adjusting and reducing the data - flow path hop count by 16% - 30%, and optimizing the error - retry - rate control mechanism by 20% - 30%.

[0124] In this embodiment, by reducing the data - acquisition frequency and the interface - call stacking - rate control mechanism, the system load can be effectively reduced, the latency can be decreased, and the stability of the system can be improved. By enhancing the edge - computing ability, the third optimization unit ensures that the system can handle more data requests and improve the overall response ability. Reducing the path hop count and optimizing the error - retry mechanism can effectively reduce the negative impacts of latency and error retries during data transmission, ensuring smooth data transmission. The intelligent optimization module automatically identifies different performance groups based on the evaluation results and triggers corresponding optimization strategies according to the different characteristics of the groups. This automated process greatly improves the operation efficiency of the system. By adjusting different strategies, the system parameters can be finely tuned under different performance conditions, avoiding resource waste or performance degradation caused by over - adjustment. Whether it is optimizing edge - data processing, reducing path hop count, or adjusting data - acquisition frequency, the optimization strategies are aimed at improving the stability, response speed, and processing ability of the system.

[0125] The setting of the threshold value is for the convenience of comparison. Regarding the size of the threshold value, it depends on the amount of sample data and the base quantity set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameter and the quantized value is not affected.

[0126] The above - mentioned formulas are all obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value. The coefficients in the formulas are set by those skilled in the art according to the actual situation. As mentioned above, this is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent recognition system for information technology consulting, characterized in that, It includes a data acquisition module, a data analysis module, an interaction performance evaluation module, and an intelligent optimization module; The data acquisition module is used to pre-define the data flow range in the information technology consulting scenario, and collect relevant data input and output processes and interface call information to establish an interaction dependency relationship dataset; The data analysis module is used to extract features of the input and output processes, the performance characteristics of the i-th interface call and the data dependency based on the interactive dependency data set through a machine learning algorithm, establish and train a data flow model, and calculate and obtain the data acquisition frequency dynamic adjustment coefficient dttz of the i-th interface according to the acquisition frequency change, CPU share and number of errors of the i-th interface: i , the input signal jitter value ddz of the i-th interface i and the i-th interface error retry rate cwl i ; The interaction performance evaluation module is used to collect the number of hops of the data flow path transmitted by the p-th path of the i-th interface and the latency of the p-th path of the i-th interface , and correlate the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i , so as to construct the comprehensive interaction performance index Zh i ; at the same time, a performance threshold Cp is preset, which is used to generate the comparison grading result of the comprehensive interaction performance index Zh i and the performance threshold Cp; The intelligent optimization module is used to dynamically adjust the data acquisition frequency, optimize the interface call mechanism, or re-allocate the data flow path according to the comparison and grading results to improve the interaction efficiency in the information technology consulting scenario; The dynamic adjustment coefficient dttzi of the data acquisition frequency of the i-th interface is calculated through the following formula: ； Among them, represents the frequency change value obtained by subtracting the acquisition frequency of the i-th interface in the previous time period from that in the current time period, represents the average acquisition frequency of the i-th interface, represents the CPU occupancy ratio of the i-th interface, represents the average CPU occupancy ratio of the i-th interface, represents the number of errors of the i-th interface, represents the maximum tolerable number of errors of the i-th interface, which is obtained according to the fault tolerance standard predetermined by the service level agreement; 、 and represent the weight values.

2. The intelligent recognition system for information technology consulting according to claim 1, characterized in that, The data acquisition module includes an input data acquisition unit, an output data acquisition unit, and a call relationship acquisition unit; The input data acquisition unit is used to collect user interaction behaviors, record the operation path and input data, and intercept and parse the data packets returned by the third-party service; The output data acquisition unit is used to collect the external interface transmission content and data status; The call relationship acquisition unit is used to collect relevant data input and output processes and interface call information. The relevant data input and output processes include: the data flow path across modules and intermediate node information, the loss rate during the data flow process, the number of changes during the data flow process, and the number of verification times during the data flow process; The interface call information includes: the interface path record of module A calling module B, the interface name, the call order, the dependency relationship chain, the call frequency, and the call data traffic; Summarize the data collected by the input data acquisition unit, the output data acquisition unit, and the call relationship acquisition unit to establish an interaction dependency relationship dataset.

3. An intelligent recognition system for information technology consulting according to claim 1, characterized in that, The data analysis module includes a preprocessing unit, a model training unit, and a feature extraction unit; The preprocessing unit is used to remove irrelevant data, outliers, duplicate data, and perform normalization on the interaction dependency relationship dataset to make it meet the requirements of the machine learning model. Then, the model training unit uses machine learning algorithms to train the interaction dependency relationship dataset to establish a data flow model, and the feature extraction unit extracts the input signal features, output data features, data flow features, and call performance features of the i-th interface to analyze and calculate to obtain: the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i , and the error retry rate cwl of the i-th interface i ; The call performance features include the acquisition frequency, the CPU occupancy ratio, and the number of errors.

4. An intelligent recognition system for information technology consulting according to claim 3, wherein The input signal jitter value ddz of the i-th interface i is obtained as follows: S11. Collect the input signal time series data of the i-th interface, and the expression is: ; n represents the total number of points of the input signal feature. S12. Calculate the standard deviation of the input signal features of the i-th interface to obtain the signal change amplitude value : ； ； Among them, represents the input signal feature of the i-th interface at the j-th moment, represents the mean value of the input signal features of the i-th interface; S13. According to the mean value of the characteristics of the input signal of the i-th interface and the signal change amplitude value , the input signal jitter value ddz of the i-th interface is calculated and obtained through the following formula i : ； The retry rate cwl of the i-th interface error i is calculated by the following formula: ； wherein, m represents the number of error categories, including network errors, interface timeouts, and server response failure categories; Indicates the weight of each error type, Indicates the number of retry attempts of the i-th interface for the k-th error type, Indicates the total number of failures of the i-th interface for the k-th error type.

5. An intelligent recognition system for information technology consulting according to claim 1, characterized in that, The interaction performance evaluation module includes a path hop count calculation unit and a delay calculation unit for each path; The path hop count calculation unit is used to collect the number of intermediate nodes in the path during the transmission of the p-th path of the i-th interface, and calculates the data stream path hop count of the p-th path of the i-th interface through the following formula : ； Among them, is the total number of intermediate nodes in the transmission of the p-th path during the transmission process, indicates that node g is a transfer node on path p, otherwise , if the p-th path is a direct connection, that is, there is no intermediate node, then ; Each path delay calculation unit is used to calculate the delay of the p-th path of the i-th interface : ； where, is the queueing delay of the g-th node on the path, is the transmission delay of the link between node g and node g + 1.

6. An intelligent recognition system for information technology consulting according to claim 1, characterized in that The interaction performance evaluation module further includes an associated unit and an evaluation unit; The associated unit is used to obtain the hop count of the data stream path transmitted by the p-th path of the i-th interface , the delay of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the error retry rate cwl of the i-th interface i . After dimensionless processing, the comprehensive interaction performance index Zh of the i-th interface is calculated through the following formula i : ； Among them, , , , and respectively represent the number of hops of the data flow path transmitted on the p-th path of the i-th interface , the delay of the p-th path of the i-th interface , the dynamic adjustment coefficient dttz of the data acquisition frequency of the i-th interface i , the input signal jitter value ddz of the i-th interface i and the weight value of the error retry rate cwl of the i-th interface i , and .

7. An intelligent recognition system for information technology consulting according to claim 6, characterized in that, The evaluation unit is used to preset a performance threshold Cp and generate a comprehensive interaction performance index Zh for the i-th interface i and a comparison grading result with the performance threshold Cp, including: When the comprehensive interaction performance index Zh of the i-th interface i > the performance threshold Cp, it indicates that the interaction performance of the i-th interface is qualified, and a first port qualified group is generated; When the performance threshold Cp×80% ≤ the comprehensive interaction performance index Zh of the i-th interface i ≤ the performance threshold Cp, it indicates that the interaction performance of the i-th interface is unqualified, and a second port exception group is generated; When the comprehensive interaction performance index Zh of the i-th interface i < the performance threshold Cp × 80%, it indicates that the interaction performance of the i-th interface is unqualified, and a third-port exception group is generated.

8. An intelligent recognition system for information technology consulting according to claim 7, characterized in that, The intelligent optimization module includes a first optimization unit, a second optimization unit, and a third optimization unit; The first optimization unit is used to trigger the first strategy after identifying the first port qualified group, including: for the i-th port, increasing the current data acquisition frequency by 5%-10%, and enhancing the interface call stacking rate control mechanism by 10%-15%; Improving the current edge data processing rate by 5%-10%; keeping the existing path without adjusting the data flow path hop count.

9. The intelligent recognition system for information technology consulting according to claim 8, characterized in that, The second optimization unit is used to trigger the second strategy after identifying the second port abnormal group, including: for the i-th port, reducing the current data acquisition frequency by 5%-10%, reducing the interface call stacking rate control mechanism by 10%-15%; increasing the edge data processing rate by 5%-10%; adjusting and reducing the data flow path hop count by 10%-15%.

10. An intelligent recognition system for information technology consulting according to claim 8, characterized in that, The third optimization unit is used to trigger a third policy after identifying a third port anomaly group, including: for the i-th port, reducing the current data acquisition frequency by 20% - 30%, reducing the interface call stack rate control mechanism by 16% - 30%; increasing the edge data processing rate by 11% - 20%; adjusting and reducing the number of hops in the data flow path by 16% - 30%, and optimizing the error retry rate control mechanism by 20% - 30%.

Images