Information technology service system based on technical service data analysis

Abstract

The application relates to the technical field of information technology service, and discloses an information technology service system based on technical service data analysis, which comprises a service data sensing module, a service value evaluation module, a service strategy generation module and a service hub management module; the system further comprises a self-adaptive learning and optimization engine; the service data sensing module collects and structurally records multidimensional dynamic data such as technical interaction, resource consumption and user feedback in the whole life cycle of a project, constructs a service process data chain that can be associated, and carries out multidimensional quantitative analysis on the data chain in combination with the service value evaluation module, so that the evaluation and deep insight into the value of technical service are realized, and the efficiency bottleneck, potential risk and optimization opportunity point in the service process are identified; the service strategy generation module automatically generates resource scheduling, technical scheme optimization and process improvement strategies based on the quantitative results of the value evaluation, so that the efficiency and effect of service execution are improved.

Description

Technical Field

[0001] This invention relates to the field of information technology service technology, specifically to an information technology service system based on technical service data analysis. Background Technology

[0002] Data analysis refers to the process of analyzing a large amount of collected data using appropriate statistical analysis methods, summarizing, understanding, and digesting the data in order to maximize its functionality and role. Data analysis is the process of studying and summarizing data in detail in order to extract useful information and form conclusions.

[0003] Currently, in the field of information technology services, the management and optimization of service processes often rely on scattered and localized data and experience-based judgments. Due to the lack of ability to systematically collect and correlate multi-dimensional dynamic data such as technical interactions, resource consumption, and user feedback throughout the entire service lifecycle, it is difficult to build a complete service process data chain.

[0004] Therefore, an information technology service system based on technical service data analysis is proposed to solve the above problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an information technology service system based on technical service data analysis, which solves the problem mentioned in the background that the system lacks the ability to systematically collect and correlate multi-dimensional dynamic data such as technical interactions, resource consumption, and user feedback throughout the entire service lifecycle, making it difficult to construct a complete service process data chain.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an information technology service system based on technical service data analysis, the system comprising a service data perception module, a service value assessment module, a service strategy generation module, and a service central management module; The service data perception module is used to collect and structure and record multi-dimensional dynamic service data throughout the entire lifecycle of a technical service project. The dynamic service data includes technical interaction data, resource consumption data, and user feedback data, and constructs an associative service process data chain. The service value assessment module is used to perform multi-dimensional quantitative analysis on the service process data chain. It calculates the quality indicators, efficiency indicators, cost indicators and comprehensive value indicators of the technical service through a pre-set assessment model, and identifies the efficiency bottlenecks, potential risks and optimization opportunities in the service process. The service strategy generation module is used to automatically generate, optimize and adjust specific technical service execution strategies based on the quantitative evaluation results output by the service value evaluation module and in combination with preset business rules and objectives. The execution strategies include resource scheduling strategies, technical solution optimization strategies and service process improvement strategies. The service hub management module is used to coordinate and control the collaborative workflow of the service data perception module, service value assessment module and service strategy generation module, to uniformly collect, govern and store data of the entire system, and to provide a multi-role interactive interface for service managers, technical implementers and customers, supporting the visual configuration of service solutions, real-time monitoring of service processes and data report presentation of service decisions.

[0007] Preferably, the service data perception module includes an interaction perception unit, a resource monitoring unit, and a feedback collection unit; The interactive sensing unit is used to capture logs, event sequences, and state change data of human-computer interaction and communication between technical components during the technical service process; The resource monitoring unit monitors and records in real time the usage and performance indicators of computing resources, storage resources, network bandwidth, and third-party service resources consumed by the execution of technical services. The feedback collection unit collects multi-source subjective evaluation and objective effect feedback data from service requesters, service implementers, and related third parties through structured forms, conversation record analysis, and sentiment analysis.

[0008] Preferably, the interactive sensing unit is deployed at the microservice gateway entry point of the technical service toolchain and at the application layer data exchange interface of the core business server, and collects operation sequences and system event streams through non-intrusive probe technology; The resource monitoring unit supports integration with mainstream cloud resource monitoring platforms, container orchestration platforms, and application performance management tools, enabling unified retrieval and standardized resource data in a hybrid environment. The feedback acquisition unit integrates a natural language processing component for performing sentiment polarity analysis, key request extraction, and problem classification on unstructured text feedback.

[0009] Preferably, the service value assessment module includes a data preprocessing unit, an indicator calculation unit, and an insight analysis unit; The data preprocessing unit is used to clean, denoise, normalize, and perform feature engineering on the raw service data to provide data input for evaluation. The indicator calculation unit has a built-in configurable evaluation indicator library and calculation engine, which periodically calculates the quantitative evaluation indicators of various services based on the preset evaluation model. The insight analysis unit uses correlation analysis, trend prediction, and root cause localization algorithms to conduct in-depth analysis of the calculated evaluation indicators and automatically identify the correlation between abnormal service patterns, efficiency trends, and service value.

[0010] Preferably, the indicator calculation unit adopts a weighted comprehensive evaluation model, and its implementation process is as follows: First, select several basic evaluation indicators related to the current evaluation objectives from the evaluation indicator library; Secondly, each selected basic evaluation metric is assigned a dynamic weight coefficient based on its importance in the current service context; Then, the original measurement values ​​of each basic evaluation indicator are standardized to eliminate the influence of dimensions; Finally, the standardized index values ​​are multiplied by their corresponding dynamic weight coefficients and summed to obtain the final quantitative evaluation score of the comprehensive service value.

[0011] The insight analysis unit uses time-series-based pattern recognition and clustering analysis methods to compare the changes in multi-dimensional indicators within the current service cycle with the pattern characteristics of historical successful service case libraries and failed service case libraries to assess the health and risk level of the current service path.

[0012] Preferably, the service policy generation module includes a policy reasoning unit, a policy verification unit, and a policy deployment unit; Based on the evaluation conclusions and optimization objectives output by the service value evaluation module, the strategy reasoning unit uses a reinforcement learning model to derive a preliminary set of service optimization strategies. The strategy verification unit uses historical data playback to simulate and extrapolate the preliminary strategy set, predicts the impact of its execution on key service value indicators, and selects the effective strategy subset with the highest expected return. The strategy deployment unit converts the verified optimal service optimization strategy into specific technical service toolchain executable instructions, configuration changes, and workflow adjustment schemes, and pushes them to the corresponding service execution environment.

[0013] Preferably, the strategy reasoning unit integrates an association rule learning algorithm to mine high-frequency co-occurring high-quality service behavior combinations from historical successful service cases and solidify them into reusable strategy rules; The strategy verification unit constructs a lightweight digital twin sandbox environment, imports the current service's technology stack and configuration, and executes and verifies the candidate strategies in the sandbox. The strategy deployment unit interfaces with the technical service management platform, the continuous integration and continuous deployment pipeline, and the automated operation and maintenance platform, enabling the optimization strategy to be applied to the target service environment in a one-click and progressive manner.

[0014] Preferably, the service hub management module includes a data hub unit, a process orchestration unit, and a human-computer interaction unit; The data hub unit serves as a unified data warehouse for the entire system, standardizing, integrating, and persistently storing data from various modules, and providing a unified data service interface for each module to call. The process orchestration unit defines and drives a closed-loop workflow of data perception, value assessment, strategy generation, strategy execution, and effect feedback, and triggers collaborative operations of each module according to preset conditions. The human-computer interaction unit provides differentiated operation views and data cockpits for different roles, supports drag-and-drop construction of service panoramas, visual configuration of evaluation models, logical editing of strategy rules, and multi-dimensional chart display of analysis results.

[0015] Preferably, the data hub unit adopts a data lake warehouse integrated architecture, which supports low-cost storage and exploratory analysis of multi-source heterogeneous raw data, and models the cleaned data according to subject domains to form dimensional models for ad-hoc query and report generation. The process orchestration unit is implemented using a low-code process engine, which allows system administrators to customize complex data analysis pipelines, cross-module business rules, and automated triggering conditions through a graphical interface. The human-computer interaction unit has a built-in collaboration function, which supports online annotation, commenting and task assignment within the service team around service evaluation reports and optimization strategies, forming a data-based collaborative decision-making mechanism.

[0016] Preferably, the system further includes an adaptive learning and optimization engine, which continuously monitors the actual execution effect of the deployed service optimization strategy and uses the new service data after execution as feedback input to dynamically adjust the evaluation model parameters in the service value evaluation module and optimize the strategy reasoning rules in the service strategy generation module.

[0017] Compared with existing technologies, the present invention provides an information technology service system based on technical service data analysis, which has the following beneficial effects: 1. In this invention, the service data perception module collects and structurally records multi-dimensional dynamic data such as technical interactions, resource consumption, and user feedback throughout the entire project lifecycle, constructs a service process data chain that can be linked, and combines the service value assessment module to conduct multi-dimensional quantitative analysis of the data chain, thereby achieving the assessment and in-depth insight into the value of technical services and identifying efficiency bottlenecks, potential risks, and optimization opportunities in the service process.

[0018] 2. In this invention, the service strategy generation module automatically generates resource scheduling, technical solution optimization, and process improvement strategies based on the quantitative results of value assessment. It also uses historical data playback and digital twin sandbox for simulation verification and screening. Finally, the optimal strategy is deployed to the execution environment, enabling the system to generate, verify, and apply service strategies in a closed loop based on data-driven approaches, thereby improving the efficiency and effectiveness of service execution.

[0019] 3. In this invention, the service hub management module performs unified data aggregation and governance, coordinates and controls the workflow of data perception, value assessment, strategy generation and execution, and provides interactive interfaces and data dashboards for multiple roles. At the same time, it continuously monitors the effect of strategies and provides feedback on optimization models and rules through adaptive learning and optimization engines, enabling the system to achieve cross-role and cross-process collaborative management and decision support. Attached Figure Description

[0020] Figure 1 This is an architecture diagram of an information technology service system based on technical service data analysis according to the present invention; Figure 2 This is a unit architecture diagram of the service value assessment module in this invention; Figure 3 This is a flowchart illustrating the operational steps of an information technology service system based on technical service data analysis, as described in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] For specific implementation examples, please refer to: Figures 1-3 An information technology service system based on technical service data analysis, comprising a service data perception module, a service value assessment module, a service strategy generation module, and a service central management module; The service data awareness module is used to collect and structure and record multi-dimensional dynamic service data throughout the entire lifecycle of a technical service project. The dynamic service data includes technical interaction data, resource consumption data, and user feedback data, and constructs a service process data chain that can be linked. The service value assessment module is used to perform multi-dimensional quantitative analysis of the service process data chain. Through the pre-set assessment model, it calculates the quality indicators, efficiency indicators, cost indicators and comprehensive value indicators of technical services, and identifies the efficiency bottlenecks, potential risks and optimization opportunities in the service process. The service strategy generation module is used to automatically generate, optimize and adjust specific technical service execution strategies based on the quantitative evaluation results output by the service value assessment module and in combination with preset business rules and objectives. The execution strategies include resource scheduling strategies, technical solution optimization strategies and service process improvement strategies. The service hub management module coordinates and controls the collaborative workflow of the service data perception module, service value assessment module, and service strategy generation module. It unifies the aggregation, governance, and storage of data across the entire system and provides a multi-role interactive interface for service managers, technical implementers, and customers. It supports the visual configuration of service solutions, real-time monitoring of service processes, and the presentation of data reports for service decisions.

[0023] The service data perception module includes an interaction perception unit, a resource monitoring unit, and a feedback collection unit; The interactive awareness unit is used to capture logs, event sequences, and state change data of human-computer interaction and communication between technical components during the technical service process. Specifically, in implementation: By deploying traffic probes at the operating system kernel level, remote procedure call messages, database query statements, and underlying socket communication data packets between various software components and microservices during the technical service process can be intercepted non-intrusively, and structured interaction logs and event sequences can be extracted from them. By using automated testing frameworks or front-end script injection techniques, we can monitor the focus changes and operation behaviors of graphical user interface elements on user terminals in real time, capture user clicks, inputs and page navigation commands, and form human-computer interaction logs. The monitoring probe is injected into the running process in the form of a dynamic link library. Aspect-oriented programming technology is used to intercept the execution entry and exit of key business functions, thereby sensing and recording the system's global state machine transition data and abnormal events. The resource monitoring unit monitors and records in real time the usage and performance indicators of computing resources, storage resources, network bandwidth, and third-party service resources consumed by the execution of technical services. The feedback collection unit collects multi-source subjective evaluations and objective effect feedback data from service requesters, service implementers, and related third parties through structured forms, conversation record analysis, and sentiment analysis.

[0024] The interactive perception unit is deployed at the microservice gateway entry point of the technical service toolchain and at the application layer data exchange interface of the core business server. It collects operation sequences and system event streams through non-intrusive probe technology. The resource monitoring unit supports integration with mainstream cloud resource monitoring platforms, container orchestration platforms, and application performance management tools, enabling unified retrieval and standardized resource data in hybrid environments. The standardization formula is: ; in The result is after standardization and its value range is [0,1]. These are the original values. , These are the normalized upper and lower limit benchmark values; The feedback collection unit integrates natural language processing components for performing sentiment polarity analysis, key request extraction, and problem classification on unstructured text feedback; After receiving the raw data, the feedback collection unit first calls the unified pull interface, which uses the adapter pattern to be compatible with the data protocols of different technology service management platforms, i.e., RESTful API, to achieve centralized data retrieval across platforms; secondly, the system executes a standardized processing procedure for the retrieved multi-source heterogeneous data. For unstructured text feedback, the natural language processing component employs a multi-task learning model. This model includes a decoder for sentiment polarity analysis, outputting sentiment scores; a key appeal extraction module based on bidirectional encoder representations, identifying core entities in the text; and a conditional random field-based question classification layer, assigning predefined business labels to the feedback. In practice, the system inputs standardized text features into the classifier, outputting probability distributions for each category. The calculation formula is as follows: ; in This represents the probability distribution vector of the input text after the natural language processing component processes its features, distributed across various predefined question categories. This represents the hyperbolic tangent activation function. , This represents the parameter matrix in the problem classification model. , This represents the bias vector in the question classification model. Using this formula, the system can automatically convert unstructured text into standard structured data with sentiment polarity, key demands, and question classification for use by subsequent modules.

[0025] The service value assessment module includes a data preprocessing unit, an indicator calculation unit, and an insight analysis unit; The data preprocessing unit is used to clean, denoise, normalize, and perform feature engineering on the raw service data to provide data input for evaluation; First, data cleaning is performed using the 3σ criterion and KNN algorithm to handle anomalies and missing data. Then, standardization is used to standardize different indicators to a unified range. Finally, feature engineering is performed by calculating moving averages and constructing sine-cosine periodic functions, thereby transforming the raw service monitoring data into feature vectors that can be directly used by the indicator calculation unit and the insight analysis unit. Outlier removal: ; in For at a certain point in time The original observations, The arithmetic mean of all observations. The standard deviation of all observations; The K-nearest neighbor algorithm is used for imputation. Its core principle is to find the K most similar known samples to the missing sample and then take a weighted average of their values. Missing value imputation: ; ; ; in This represents the imputed estimate of the missing value. This represents the feature vector of the sample to be imputed, which contains missing values. Indicates that in the known complete dataset, with Most similar The feature vectors of the neighboring samples Indicates the sample to be interpolated and the first... The feature vectors of the neighboring samples , Representing vectors respectively and The 1 eigenvalue, Represents the total dimension of the features. This represents the preset number of nearest neighbor samples. Indicates based on distance The reciprocal of the first digit is used to calculate the second digit. The weights of neighboring samples are assigned, with greater weights for closer samples. Extracting moving average features: ; in Indicates at a point in time Calculated moving average, The window size representing the moving average. Indicates at a point in time The previous The index observation value at each moment; Extracting periodic fluctuation features: ; ; in , This indicates the periodic fluctuation characteristics of the structure. Indicates a time index. This indicates the preset period length. Represents pi; The indicator calculation unit has a built-in configurable evaluation indicator library and calculation engine, which periodically calculates quantitative evaluation indicators for various services based on preset evaluation models. The insight and analysis unit uses correlation analysis, trend prediction and root cause localization algorithms to conduct in-depth analysis of the calculated evaluation indicators and automatically identify abnormal service patterns, performance trends and the relationship between service value. When the insight analysis unit initiates the deep analysis algorithm, in order to automatically identify abnormal service patterns, the system uses the Isolation Forest algorithm to scan multi-dimensional indicator data, calculate the path length of each sample, and then obtain an anomaly score for evaluation. ; in Indicates sample In a forest composed of many isolated trees The more likely the sample is to be an outlier, the higher the score is to be 1. This indicates the sample data points that need to be evaluated. This represents a constructed forest of isolated trees. Indicates sample In the forest Path lengths on all isolated trees The average value, This refers to constructing a forest given a certain number of samples. The average path length baseline is used to standardize path length.

[0026] The indicator calculation unit adopts a weighted comprehensive evaluation model, and its implementation process is as follows: First, select several basic evaluation indicators related to the current evaluation objectives from the evaluation indicator library; Secondly, assign dynamic weight coefficients to each selected basic evaluation metric based on its importance in the current service context: ; in Indicates the first The final dynamic weight coefficients of each evaluation indicator are determined after context matching. Indicates the first The initial static weights of each indicator are preset in the basic evaluation indicator library. Indicates the first The feature vectors of each indicator and the feature vector of the current service context Semantic similarity between them Indicates the adjustment parameter and usually satisfies ; Then, the original measurement values ​​of each basic evaluation indicator are standardized to eliminate the influence of dimensions; ; in Indicates the first The standardized values ​​of the basic evaluation indicators Indicates the number obtained from the monitoring system The original measured values ​​of each indicator, , These represent the first time in the historical dataset or the current evaluation period, respectively. The maximum and minimum values ​​observed for each indicator; Finally, the standardized index values ​​are multiplied by their corresponding dynamic weight coefficients and summed to obtain the final quantitative evaluation score of the comprehensive service value: ; in This represents the final quantitative assessment score of the overall service value. A higher score indicates a healthier service status, while a lower score suggests potential risks. This indicates the total number of basic evaluation indicators included in the calculation of the current evaluation model; The insight analysis unit uses time-series-based pattern recognition and clustering analysis methods to compare the changes in multi-dimensional indicators within the current service cycle with the pattern characteristics of historical successful service case libraries and failed service case libraries to assess the health and risk level of the current service path. The system performs cluster analysis on all multivariate time series of historical service cases. Using the k-means clustering algorithm, it divides the massive historical sequences into several clusters with common pattern characteristics. The center of each cluster is not the arithmetic mean, but is obtained by calculating the DTW centroid of all sequences within the cluster. The algorithm iteratively updates the cluster centers and cluster assignments, aiming to minimize the sum of the DTW distances of all sequences to their respective cluster centers. Finally, the system constructs two core pattern libraries: a successful case pattern library and a failed case pattern library. The system calculates the DTW distance between the current service cycle's time series and each representative pattern in the pattern library. Then, it identifies the K nearest neighbor patterns with the smallest distance to the current sequence and records the number of these nearest neighbor patterns from the successful case library and the failed case library, respectively. Simultaneously, the system extracts the trend characteristics of the current sequence itself to determine whether the indicators are improving or deteriorating. Finally, the system uses a weighted scoring model to quantitatively evaluate the health of the current service path. ; ; in For health, , , The weighting coefficients for each evaluation dimension are as follows: , The number of samples in the successful case study pattern library. The number of samples in the failure case pattern library. This represents the total number of samples. This is a trend scoring function calculated based on the slope of the time series trend. The trend slope This is the trend sensitivity adjustment coefficient. The base of the natural logarithm is used; finally, the system maps the calculated health level to a preset risk level, namely healthy, sub-healthy, and high-risk, completing the automated assessment; when When the value falls below the threshold, a risk warning will be automatically triggered, and the strategy generation module will be notified.

[0027] The service policy generation module includes a policy reasoning unit, a policy verification unit, and a policy deployment unit; Based on the evaluation conclusions and optimization objectives output by the service value assessment module, the strategy reasoning unit uses a reinforcement learning model to derive a preliminary set of service optimization strategies, specifically including: Environment and Agent Modeling: The system abstracts the technical service environment as a Markov decision process; state Defined at time The service status vector includes metrics obtained from the value assessment module, such as resource utilization, service response time, and error rate; actions Defined as a set of executable optimization operations; reward It is a scalar signal fed back to the agent by the environment based on state changes. Its core calculation is the increment of the service value index, and the formula is: ,in For the current moment The comprehensive service value quantitative assessment score, It is a cost item used to penalize complex operations and resource consumption.

[0028] Policy learning and generation: The system uses a deep Q-network as the learning algorithm; the goal of the policy network is to learn an optimal action-value function. This function represents the state. Next action Then it follows the maximum expected cumulative discount reward obtainable by the optimal strategy; the update of the Q value follows the Bellman optimality equation: ; in Indicates the state Select action Action value function estimate, Indicates the learning rate. Indicates the state Next action Afterwards, the environment was Instant rewards with real-time feedback Represents the discount factor and , Indicates the next state The system calculates the maximum expected Q-value among all possible actions. During training, the system uses experience replay and fixed-target network techniques to stabilize the learning process. After training, for the current service state, the policy network outputs all available actions. The system selects the top Q-values. These actions constitute a preliminary set of optimization strategies; The strategy verification unit uses historical data playback to simulate and extrapolate the initial set of strategies, predicts the impact of their execution on key service value indicators, and selects the effective strategy subset with the highest expected returns. ; in For strategy Expected value To assess the number of key value indicators of interest, Strategies for simulation and prediction After execution, the first Scores of key value metrics, For the first time when the strategy is not executed Scores for key value metrics For the first The system ranks all preliminary strategies based on the calculated expected value and selects the top F strategies with positive net returns to form an effective strategy subset. This subset represents the strategy combination that can bring the maximum positive return in the simulation environment. The strategy deployment unit converts the validated optimal service optimization strategy into specific technical service toolchain executable instructions, configuration changes, and workflow adjustment plans, and pushes them to the corresponding service execution environment; After receiving the selected optimal strategy, the strategy deployment unit calls the strategy compiler. The compiler has a built-in rule library for mapping strategies and actions, which parses the abstract optimization strategy description. The generated executable solution is encapsulated into a transactional change set and pushed to the corresponding technical service management platform via API using a service account token.

[0029] The strategy reasoning unit integrates an association rule learning algorithm to mine high-frequency co-occurring high-quality service behavior combinations from historical successful service cases and solidify them into reusable strategy rules, enriching the strategy rule library; The strategy reasoning unit uses the association rule learning algorithm, namely the Apriori algorithm, to mine frequent itemsets and strong association rules in the atomic operation sequences of historical successful service cases, identifying high-frequency co-occurring combinations of high-quality service behaviors that can lead to successful results: Frequent itemset mining: Using the Apriori algorithm, all frequent itemsets that meet the minimum support threshold are discovered through a layer-by-layer search. ; in Representation Itemset Support Indicates a containing itemset The total number of all transactions, This represents the total number of historically successful transactions. Strong association rule generation: Generate all possible association rules from the mined frequent itemset, and calculate their confidence and lift to filter out meaningful strong rules; Confidence level: measures the confidence level when the combination of operations is... When it occurs, the operation combination The conditional probability, i.e., the reliability of the rule, also appears at the same time: ; in Representation rules confidence level Representation Itemset and The support of the union; Lift: In the measurement rules and The correlation between the two is such that a lift greater than 1 indicates that they are positively correlated and not independent. ; in Representation rules The degree of improvement Representation Itemset Support Representation Itemset Support level; The strategy verification unit constructs a lightweight digital twin sandbox environment, imports the current service's technology stack and configuration, executes and verifies candidate strategies within the sandbox, avoiding interference with the real service environment. The digital twin sandbox environment clones a complete technology stack image of the current service in virtual space by reading a configuration snapshot of the production environment. Before implementing a new optimization strategy, the system injects the strategy parameters into the sandbox environment, performs stress testing and effect verification without loss and with security, ensuring that the strategy has undergone rigorous simulation before actual deployment, avoiding interference with the real service environment. The strategy deployment unit interfaces with the technical service management platform, continuous integration and continuous deployment pipeline, and automated operation and maintenance platform, enabling the one-click and gradual application of optimization strategies to the target service environment.

[0030] The service hub management module includes a data hub unit, a process orchestration unit, and a human-computer interaction unit; The data hub unit serves as the unified data warehouse for the entire system. It standardizes, integrates, and persistently stores data from various modules, and provides a unified data service interface for each module to call. The process orchestration unit defines and drives a closed-loop workflow of data perception, value assessment, strategy generation, strategy execution, and effect feedback, and triggers collaborative operations of each module according to preset conditions. The human-computer interaction unit provides differentiated operation views and data cockpits for different roles, supporting drag-and-drop construction of service panoramas, visual configuration of evaluation models, logical editing of policy rules, and multi-dimensional chart display of analysis results.

[0031] The data hub unit adopts a data lake warehouse integrated architecture, which supports low-cost storage and exploratory analysis of multi-source heterogeneous raw data. The cleaned data is modeled according to subject domains to form dimensional models for ad-hoc querying and report generation. When implementing a data lake warehouse architecture, at the bottom layer, the low-cost and infinitely scalable characteristics of object storage are used to directly store raw, multi-format unstructured data; at the top layer, the Spark computing engine is used to clean and define the schema of the data, transforming it into structured data tables to support ad-hoc queries and report generation. The workflow orchestration unit is implemented using a low-code workflow engine, which allows system administrators to customize complex data analysis pipelines, cross-module business rules, and automated triggering conditions through a graphical interface, enabling flexible orchestration and agile adjustment of workflows. The human-computer interaction unit has built-in collaboration functions, which support the service team to annotate, comment on and assign tasks online around service evaluation reports and optimization strategies, forming a data-based collaborative decision-making mechanism.

[0032] The system also includes an adaptive learning and optimization engine, which continuously monitors the actual execution effect of the deployed service optimization strategies and uses the new service data after execution as feedback input to dynamically adjust the evaluation model parameters in the service value assessment module and optimize the strategy reasoning rules in the service strategy generation module. The adaptive learning and optimization engine uses an online learning mechanism to dynamically adjust the parameters of the service value assessment model by utilizing the actual execution effect data of deployed strategies, making its assessment more in line with actual business value. At the same time, it uses the reward signals obtained after strategy execution to optimize the decision function and rule base of the strategy reasoning model, improving the quality and accuracy of future strategy generation, thereby achieving the overall self-evolution of the system.

[0033] The operation steps of an information technology service system based on technical service data analysis are as follows: Step 1: Service Data Awareness Throughout the entire lifecycle of a technical service project, the service data perception module collects and structures multi-dimensional dynamic service data, including technical interaction data, resource consumption data, and user feedback data, to construct a linkable service process data chain. Specifically, this module captures technical interaction logs and event sequences through the interaction perception unit, monitors real-time resource consumption indicators such as computing, storage, and network through the resource monitoring unit, and collects multi-source subjective evaluations and objective effect feedback data through the feedback collection unit, thus providing a complete and structured data foundation for subsequent analysis.

[0034] Step Two: Service Value Assessment The service value assessment module performs multi-dimensional quantitative analysis of the service process data chain. First, the data preprocessing unit cleans, denoises, normalizes, and performs feature engineering on the raw service data. Then, the indicator calculation unit calculates the quality indicators, efficiency indicators, cost indicators, and comprehensive value indicators of the technical service periodically based on the pre-set evaluation model. Finally, the insight analysis unit uses correlation analysis, trend prediction, and root cause localization algorithms to automatically identify efficiency bottlenecks, potential risks, and optimization opportunities in the service process, and assess the health and risk level of the service path.

[0035] Step 3: Service Policy Generation Based on the quantitative evaluation results output by the service value assessment module, and combined with preset business rules and objectives, the service strategy generation module automatically generates and optimizes specific technical service execution strategies, including resource scheduling strategies, technical solution optimization strategies, and service process improvement strategies. This module first uses a strategy reasoning unit to derive a preliminary set of service optimization strategies using reinforcement learning models or association rule learning algorithms. Then, the strategy verification unit uses historical data playback or a digital twin sandbox environment to conduct simulations and selects the subset of effective strategies with the highest expected returns. Finally, the strategy deployment unit converts the verified optimal strategies into executable instructions and pushes them to the service execution environment.

[0036] Step 4: Service Hub Management The service hub management module coordinates and controls the collaborative workflow of the service data perception module, service value assessment module, and service strategy generation module. This module unifies the aggregation, governance, and storage of data across the entire system through the data hub unit, defines and drives the closed-loop workflow from data perception to effect feedback through the process orchestration unit, and provides multi-role interactive interfaces and data dashboards for service managers, technical implementers, and customers through the human-computer interaction unit. It supports the visual configuration of service solutions, real-time monitoring of service processes, and data report presentation of service decisions, realizing the integration and collaboration of all aspects of the system.

[0037] Step 5: Adaptive Learning and Optimization By continuously monitoring the actual execution effect of deployed service optimization strategies through an adaptive learning and optimization engine, and using the new service data after execution as feedback input, the evaluation model parameters in the service value assessment module and the strategy reasoning rules in the service strategy generation module are dynamically adjusted, thereby forming a continuous self-optimization closed loop based on actual service data feedback.

[0038] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An information technology service system based on technical service data analysis, characterized in that: The system includes a service data perception module, a service value assessment module, a service strategy generation module, and a service central management module. The service data perception module is used to collect and structure and record multi-dimensional dynamic service data throughout the entire lifecycle of a technical service project. The dynamic service data includes technical interaction data, resource consumption data, and user feedback data, and constructs an associative service process data chain. The service value assessment module is used to perform multi-dimensional quantitative analysis on the service process data chain. It calculates the quality indicators, efficiency indicators, cost indicators and comprehensive value indicators of the technical service through a pre-set assessment model, and identifies the efficiency bottlenecks, potential risks and optimization opportunities in the service process. The service strategy generation module is used to automatically generate, optimize and adjust specific technical service execution strategies based on the quantitative evaluation results output by the service value evaluation module and in combination with preset business rules and objectives. The execution strategies include resource scheduling strategies, technical solution optimization strategies and service process improvement strategies. The service hub management module is used to coordinate and control the collaborative workflow of the service data perception module, service value assessment module and service strategy generation module, to uniformly collect, govern and store data of the entire system, and to provide a multi-role interactive interface for service managers, technical implementers and customers, supporting the visual configuration of service solutions, real-time monitoring of service processes and data report presentation of service decisions.

2. The information technology service system based on technical service data analysis according to claim 1, characterized in that: The service data perception module includes an interactive perception unit, a resource monitoring unit, and a feedback collection unit; The interactive sensing unit is used to capture logs, event sequences, and state change data of human-computer interaction and communication between technical components during the technical service process; The resource monitoring unit monitors and records in real time the usage and performance indicators of computing resources, storage resources, network bandwidth, and third-party service resources consumed by the execution of technical services. The feedback collection unit collects multi-source subjective evaluation and objective effect feedback data from service requesters, service implementers, and related third parties through structured forms, conversation record analysis, and sentiment calculation.

3. The information technology service system based on technical service data analysis according to claim 2, characterized in that: The interactive sensing unit is deployed at the microservice gateway entry point of the technical service toolchain and at the application layer data exchange interface of the core business server, and collects operation sequences and system event streams through non-intrusive probe technology. The resource monitoring unit supports integration with mainstream cloud resource monitoring platforms, container orchestration platforms, and application performance management tools, enabling unified retrieval and standardized resource data in a hybrid environment. The feedback acquisition unit integrates a natural language processing component for performing sentiment polarity analysis, key request extraction, and problem classification on unstructured text feedback.

4. An information technology service system based on technical service data analysis according to claim 1, characterized in that: The service value assessment module includes a data preprocessing unit, an indicator calculation unit, and an insight analysis unit. The data preprocessing unit is used to clean, denoise, normalize, and perform feature engineering on the raw service data to provide data input for evaluation. The indicator calculation unit has a built-in configurable evaluation indicator library and calculation engine, which periodically calculates the quantitative evaluation indicators of various services based on the preset evaluation model. The insight analysis unit uses correlation analysis, trend prediction, and root cause localization algorithms to conduct in-depth analysis of the calculated evaluation indicators and automatically identify the correlation between abnormal service patterns, efficiency trends, and service value.

5. An information technology service system based on technical service data analysis according to claim 4, characterized in that: The indicator calculation unit adopts a weighted comprehensive evaluation model, and its implementation process is as follows: First, select several basic evaluation indicators related to the current evaluation objectives from the evaluation indicator library; Secondly, each selected basic evaluation metric is assigned a dynamic weight coefficient based on its importance in the current service context; Then, the original measurement values ​​of each basic evaluation indicator are standardized to eliminate the influence of dimensions; Finally, the standardized index values ​​are multiplied by their corresponding dynamic weight coefficients and summed to obtain the final quantitative evaluation score of the comprehensive service value. The insight analysis unit uses time-series-based pattern recognition and clustering analysis methods to compare the changes in multi-dimensional indicators within the current service cycle with the pattern characteristics of historical successful service case libraries and failed service case libraries to assess the health and risk level of the current service path.

6. An information technology service system based on technical service data analysis according to claim 1, characterized in that: The service policy generation module includes a policy reasoning unit, a policy verification unit, and a policy deployment unit; Based on the evaluation conclusions and optimization objectives output by the service value evaluation module, the strategy reasoning unit uses a reinforcement learning model to derive a preliminary set of service optimization strategies. The strategy verification unit uses historical data playback to simulate and extrapolate the preliminary strategy set, predicts the impact of its execution on key service value indicators, and selects the effective strategy subset with the highest expected return. The strategy deployment unit converts the verified optimal service optimization strategy into specific technical service toolchain executable instructions, configuration changes, and workflow adjustment schemes, and pushes them to the corresponding service execution environment.

7. An information technology service system based on technical service data analysis according to claim 6, characterized in that: The strategy reasoning unit integrates an association rule learning algorithm to mine high-frequency co-occurring high-quality service behavior combinations from historical successful service cases and solidify them into reusable strategy rules. The strategy verification unit constructs a lightweight digital twin sandbox environment, imports the current service's technology stack and configuration, and executes and verifies the candidate strategies in the sandbox. The strategy deployment unit interfaces with the technical service management platform, the continuous integration and continuous deployment pipeline, and the automated operation and maintenance platform, enabling the optimization strategy to be applied to the target service environment in a one-click and progressive manner.

8. An information technology service system based on technical service data analysis according to claim 1, characterized in that: The service hub management module includes a data hub unit, a process orchestration unit, and a human-computer interaction unit; The data hub unit serves as a unified data warehouse for the entire system, standardizing, integrating, and persistently storing data from various modules, and providing a unified data service interface for each module to call. The process orchestration unit defines and drives a closed-loop workflow of data perception, value assessment, strategy generation, strategy execution, and effect feedback, and triggers collaborative operations of each module according to preset conditions. The human-computer interaction unit provides differentiated operation views and data cockpits for different roles, supports drag-and-drop construction of service panoramas, visual configuration of evaluation models, logical editing of strategy rules, and multi-dimensional chart display of analysis results.

9. An information technology service system based on technical service data analysis according to claim 8, characterized in that: The data hub unit adopts a data lake warehouse integrated architecture, which supports low-cost storage and exploratory analysis of multi-source heterogeneous raw data. The cleaned data is modeled according to subject domains to form a dimensional model for ad-hoc querying and report generation. The process orchestration unit is implemented using a low-code process engine, which allows system administrators to customize complex data analysis pipelines, cross-module business rules, and automated triggering conditions through a graphical interface. The human-computer interaction unit has a built-in collaboration function, which supports online annotation, commenting and task assignment within the service team around service evaluation reports and optimization strategies, forming a data-based collaborative decision-making mechanism.

10. An information technology service system based on technical service data analysis according to claim 1, characterized in that: The system also includes an adaptive learning and optimization engine, which continuously monitors the actual execution effect of the deployed service optimization strategies and uses the new service data after execution as feedback input to dynamically adjust the evaluation model parameters in the service value assessment module and optimize the strategy reasoning rules in the service strategy generation module.

Images