Quality evaluation method, device, equipment, medium and product

Abstract

The invention relates to a quality evaluation method and device, equipment, a medium and a product. The method comprises the steps of obtaining to-be-evaluated service data; for each business service, performing feature extraction on the to-be-evaluated service data corresponding to the business service to obtain service feature data; and determining a target quality score of the business service according to the service feature data and the target evaluation weights corresponding to the plurality of evaluation indexes under the business service. The target evaluation weight corresponding to each evaluation index is configured for different business services, so that the calculated target quality score meets the evaluation requirement of the corresponding business service, the accuracy of the target quality score is improved, the service resources of the power service system are scheduled according to each business service, and the service quality of the power service system is improved. Balancing among service resources of all business services of the electric power service system is improved, the overall quality of the electric power service system is improved, and the comprehensive experience of a user is improved.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a quality evaluation method, apparatus, equipment, medium, and product. Background Technology

[0002] In the operation and maintenance scenarios of power service systems, the assessment of customer service quality mainly relies on post-service follow-up questionnaires, manual rating records, or standardized satisfaction / dissatisfaction labels. These traditional practices typically use a uniform question template and score service quality based on immediate customer feedback after the service is completed.

[0003] However, with the increasing personalization and diversification of customer expectations and the growing complexity of service scenarios, traditional evaluation methods have gradually exposed various problems, leading to low accuracy in evaluation results. This limits the power service system's ability to optimize the allocation of service resources. Therefore, the service quality evaluation of the power service system has become a key weakness in building an intelligent service operation system.

[0004] Against this backdrop, there is an urgent need for a service quality evaluation scheme with high accuracy to support the power service system in optimizing the allocation of service resources. Summary of the Invention

[0005] Therefore, it is necessary to provide a quality evaluation method, apparatus, equipment, medium, and product to address the aforementioned technical problems, thereby improving the accuracy of service quality evaluation and supporting the power service system in optimizing the allocation of service resources.

[0006] Firstly, this application provides a quality evaluation method, including:

[0007] Acquire service data to be evaluated for different business services within a preset time period from the power service system; different business services are implemented using different business resources;

[0008] For each business service, feature extraction is performed on the service data to be evaluated corresponding to the business service to obtain the corresponding service feature data.

[0009] Based on service characteristic data and the target evaluation weights corresponding to multiple evaluation indicators under the business service, the target quality score of the business service is determined; among them, the target quality score of each business service is used to schedule the service resources of the power service system.

[0010] In one embodiment, the target evaluation weights are determined as follows: historical service data and rating labels for different business services of the power service system within a historical time period are obtained; a fitness function is constructed based on the historical service data and rating labels, with the evaluation weights as independent variables and the goal of minimizing the rating error; and a genetic algorithm is used to determine the evaluation weights for each target based on the fitness function.

[0011] In one embodiment, a fitness function is constructed based on historical service data and rating labels, with evaluation weights as independent variables and minimizing rating error as the objective. This includes: determining a first historical quality score for the business service based on historical service data and a preset service quality evaluation model; and constructing a fitness function based on the first historical quality score, rating labels, historical service data, and initial values ​​of each evaluation weight.

[0012] In one embodiment, determining the target quality score of a business service based on service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service includes: determining a first quality score of the business service based on service feature data and a preset service quality evaluation model; determining a second quality score of the business service based on service feature data and the target evaluation weights corresponding to each evaluation indicator; and determining the target quality score of the business service based on the first quality score and the second quality score.

[0013] In one embodiment, determining the second quality score of a business service based on service feature data and the target evaluation weights corresponding to each evaluation indicator includes: multiplying each feature value in the service feature data by the target evaluation weight of the corresponding evaluation indicator in each evaluation indicator to obtain the multiplication result corresponding to each feature value; and summing the multiplication results to obtain the second quality score.

[0014] In one embodiment, feature extraction of the service data to be evaluated to obtain service feature data includes: standardizing the structured data in the service data to be evaluated to obtain first feature data; encoding the unstructured data in the service data to be evaluated to obtain second feature data; and concatenating and fusing the first feature data and the second feature data to obtain service feature data.

[0015] Secondly, this application also provides a quality evaluation device, comprising:

[0016] The data acquisition module is used to acquire service data to be evaluated for different business services in the power service system within a preset time period; different business services are implemented using different business resources.

[0017] The feature extraction module is used to extract features from the service data to be evaluated for each business service, and obtain the corresponding service feature data.

[0018] The scoring determination module is used to determine the target quality score of a business service based on service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service; among which, the target quality score of each business service is used to schedule the service resources of the power service system.

[0019] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method provided in the first aspect.

[0020] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in the first aspect.

[0021] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the method provided in the first aspect.

[0022] The aforementioned quality evaluation methods, devices, equipment, media, and products extract features from the service data to be evaluated for each business service in the power service system, obtaining service feature data corresponding to that business service. Then, based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service, a target quality score for the business service is determined. Because target evaluation weights are configured for each evaluation indicator for different business services, the calculated target quality score meets the evaluation requirements of the corresponding business service, improving the accuracy of the target quality score. This facilitates the scheduling of service resources in the power service system according to each business service, improving the balance of service resources among the various business services in the power service system, enhancing the overall quality of the power service system, and improving the overall user experience. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a quality evaluation method in one embodiment;

[0025] Figure 2 This is a flowchart illustrating the feature extraction steps in one embodiment;

[0026] Figure 3 A flowchart illustrating the target quality score determination steps in one embodiment;

[0027] Figure 4 This is a flowchart illustrating the second quality score determination step in one embodiment;

[0028] Figure 5This is a flowchart illustrating the steps for determining the target evaluation weights in one embodiment.

[0029] Figure 6 This is a flowchart illustrating the fitness function construction steps in one embodiment;

[0030] Figure 7 This is a structural block diagram of a quality evaluation device in one embodiment.

[0031] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0033] In one exemplary embodiment, a quality assessment method is provided. See also Figure 1 Quality evaluation methods include:

[0034] S110: Obtain the service data to be evaluated corresponding to different business services in the power service system within a preset time period; different business services are implemented using different business resources.

[0035] The preset time period can be set as needed, such as within the last month or the last week, or other time periods, which are not limited here.

[0036] The business services may include work order services (e.g., payment services) and customer feedback services, and may also include other business services, which are not limited here.

[0037] The data source for the service data to be evaluated can be at least one of the customer service management platform, work order system, and customer service feedback platform in the power service system.

[0038] The service data to be evaluated differs for different business services. For work order services, the service data to be evaluated may include at least one of the following: service type, response time, processing time, customer rating, and historical satisfaction rate; this type of data is structured data. For customer feedback services, the service data to be evaluated may include at least one of the following: customer reviews, complaint content, and service communication records; this type of data is unstructured data.

[0039] Different business services require different business resources to be implemented. There are various types of business resources, such as server resources (e.g., the number and types of servers required to implement the corresponding business service), storage resources, or computing resources, and of course, other business resources as well, which are not limited here.

[0040] S120: For each business service, extract features from the service data to be evaluated corresponding to the business service to obtain the corresponding service feature data.

[0041] There are various methods for feature extraction, and different methods can be used for feature extraction for different types of data.

[0042] In one alternative implementation, see Figure 2 The feature extraction steps in S120 include:

[0043] S210, standardize the structured data in the service data to be evaluated to obtain the first feature data.

[0044] Structured data includes two categories: categorical data and numerical data. For categorical data, standardization methods can include, but are not limited to, one-hot encoding. For numerical data, standardization methods can include, but are not limited to, using min-max normalization to map numerical values ​​and eliminate scale differences.

[0045] For example, categorical data may include service type, service channel, or customer level, and is quantified using one-hot encoding. One-hot encoding maps each category to a dimension, where "1" indicates that the category is active, and all other positions are set to "0," thus avoiding pseudo-order relationships based on numerical values ​​between categories. For instance, if the service type field has three categories: "Fault Reporting," "Power Outage Inquiry," and "Service Processing," then the field data for the service type field is encoded as a three-dimensional vector:

[0046]

[0047] For example, for numerical data, the following formula can be used for standardization:

[0048]

[0049] In the formula, This represents the result after standardization. It is the minimum value in the numerical data. Let x be the maximum value in the numerical data, where x is the numerical data.

[0050] S220, the unstructured data in the service data to be evaluated is encoded to obtain the second feature data.

[0051] For unstructured data (such as customer review texts and complaint statements), the TF-IDF (Term Frequency-Inverse Document Frequency) method can be used to transform the unstructured data into a vector representation, fully capturing the keyword density and sentiment in customer subjective feedback. Specifically, the unstructured data is first segmented into Chinese words, then a bag-of-words model is constructed, and the importance of each term in the unstructured data is calculated based on term frequency and inverse document frequency. Finally, a sparse feature vector T of length d is generated, which reflects the keyword weight distribution in the unstructured data and can be used to characterize customer subjective emotions and service content priorities.

[0052] S230, the first feature data and the second feature data are spliced ​​and fused to obtain service feature data.

[0053] For example, if the first feature data corresponding to the structured data is S and the second feature data is T, then the service feature data X = [S, T].

[0054] The above implementation method extracts features from both structured and unstructured data of the service data to be evaluated, and then concatenates and merges the two types of feature data to obtain service feature data. This makes the extracted service feature data vectorized, normalized, and consistent, effectively supporting the high-dimensional input requirements of subsequent neural network modeling. Based on this implementation method, semantic information parsing can improve the accuracy of subsequent scoring when dealing with customer feedback text, customer service records, complaint logs, and other information.

[0055] S130, based on service characteristic data and the target evaluation weights corresponding to multiple evaluation indicators under the business service, determine the target quality score of the business service; among which, the target quality score of each business service is used to schedule the service resources of the power service system.

[0056] Among them, different evaluation indicators can be understood as different dimensions for evaluating business services.

[0057] The target evaluation weight of the evaluation indicator can be understood as the importance of the evaluation result obtained by using the evaluation indicator to evaluate the business service; the higher the importance, the higher the target evaluation weight.

[0058] The target quality score can be understood as the comprehensive quality score.

[0059] Among them, based on the service characteristic data and the target evaluation weights corresponding to multiple evaluation indicators, various methods can be used to determine the target quality score, which are not limited here.

[0060] Among them, the target evaluation weights corresponding to multiple evaluation indicators are weights set for business services.

[0061] Understandably, after obtaining the target quality score of a service, the service resources occupied by that service can be allocated based on that score. For example, if the target quality score of service A is much higher than that of service B, some of the service resources currently occupied by service A can be allocated to service B to improve service B's quality score, thereby improving the overall quality of the power service system.

[0062] The aforementioned quality evaluation method extracts features from the service data to be evaluated for each business service in the power service system, obtaining service feature data corresponding to that business service. Then, based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service, a target quality score for the business service is determined. Because target evaluation weights are configured for each evaluation indicator for different business services, the calculated target quality score meets the evaluation requirements of the corresponding business service, improving the accuracy of the target quality score. This facilitates the scheduling of service resources in the power service system according to each business service, improving the balance of service resources among the various business services in the power service system, enhancing the overall quality of the power service system, and improving the overall user experience.

[0063] Based on the technical solutions provided in the above embodiments, an optional embodiment is provided, in which the target quality score determination step in S130 is refined.

[0064] See Figure 3 The steps for determining the target quality score include:

[0065] S310, based on service characteristic data and a preset service quality evaluation model, determines the first quality score of the business service.

[0066] That is, the service feature data is input into the preset service quality scoring model to obtain the first quality score.

[0067] The service quality evaluation model pre-learns the non-linear relationship between the service process and customer satisfaction results. It possesses strong fitting and generalization capabilities, adapting to the multi-dimensional, high-noise, and weakly linearly correlated characteristics of service data, enabling intelligent prediction of service quality scores.

[0068] The service quality evaluation model can be a multilayer perceptron neural network (MLN) model or other models, without limitation. The MNN model can include an input layer, at least one hidden layer, and an output layer. The nodes in each layer are organized in a fully connected manner, and the activation function can be the sigmoid function to enhance nonlinear expressiveness. The basic structure of the MNN model is shown in the figure below.

[0069] For the input layer, the input service feature data is: The input layer passes it to the hidden layer for the first nonlinear transformation. The output of each hidden layer node is denoted as . The calculation formula is as follows:

[0070]

[0071] In the formula, Represents the first input layer The node and the first hidden layer The connection weights between nodes For the hidden layer The bias term of each node, Let be the i-th feature value in the service feature data, and n be the number of feature values ​​in the service feature data. The activation function is... The Sigmoid function is selected and defined as follows:

[0072]

[0073] Of course, the above process can also be extended to multi-layered hidden structures.

[0074] For the output layer, the number of nodes in the output layer is Then the output layer's first The formula for calculating the output of each node is as follows:

[0075]

[0076] in, This represents the connection weight between the j-th node in the hidden layer and the k-th node in the output layer. For the bias of the output layer, activation function We also use the Sigmoid function.

[0077] Understandably, during the training process of a multilayer perceptron neural network model, let the expected output of each node in the output layer be... The prediction error function E is then defined as the squared loss form:

[0078]

[0079] The standard backpropagation algorithm combined with gradient descent is used to update the weight parameters in the multilayer perceptron model in order to minimize the error function. Among them, the adjustment amount of the weight. Adjustment amount with bias The determination method is as follows:

[0080]

[0081] in, and The learning rate parameter, This is the gradient factor for the error term of the output layer.

[0082] Through the aforementioned forward and backward propagation processes, the multilayer perceptron model can automatically learn the complex mapping relationship between service features and satisfaction scores. After training, inputting any set of service feature vectors will output the corresponding predicted score, providing the objective function input for subsequent optimization modules.

[0083] S320 determines the second quality score of the business service based on the service feature data and the target evaluation weights corresponding to each evaluation indicator.

[0084] The process of determining the second quality score reflects the linear relationship between the business service process and the customer satisfaction result.

[0085] The target evaluation weights corresponding to the evaluation indicators reflect the comprehensive importance and dynamic contribution of the evaluation indicators in the current business scenario.

[0086] In one alternative implementation, see Figure 4 The steps for determining the second quality score in S320 include:

[0087] S410, multiply each feature value in the service feature data by the target evaluation weight of the corresponding evaluation indicator in each evaluation indicator to obtain the multiplication result corresponding to each feature value.

[0088] S420: The results of each multiplication are summed to obtain the second quality score.

[0089] Based on the steps above, the second quality score can be calculated using the following formula:

[0090]

[0091] In the formula, This is the second quality score, where N is the number of evaluation indicators. Let i be the evaluation weight corresponding to the i-th evaluation indicator. This refers to the i-th feature value in the service feature data.

[0092] In the above implementation, the second quality score is obtained by summing the products of each feature value in the service feature data and the target evaluation weight of the corresponding evaluation indicator in each evaluation indicator. It can be seen that the calculation process of the second quality score can reflect the linear relationship between the process of business service to customer service and the satisfaction result.

[0093] S330, determine the target quality score for the business service based on the first quality score and the second quality score.

[0094] The target quality score can be obtained by weighted summation of the first and second quality scores, as shown in the following expression:

[0095]

[0096] In the formula, Score the target quality. It received the highest quality rating. As the second quality score, The weight of the first quality score. The weights for the second quality score. The value range is [0,1].

[0097] in, It can be set according to actual business preferences, for example, when emphasizing model evaluation, take When emphasizing human interpretability, take This fusion approach maintains predictive capabilities while introducing explicit indicator weighting, thereby enhancing the business transparency of the scoring results.

[0098] In this embodiment, the process of determining the first quality score can reflect the non-linear relationship between the process of business service to customer service and the satisfaction result, the process of determining the second quality score can reflect the linear relationship between the process of business service to customer service and the satisfaction result, and the process of determining the target quality score can comprehensively reflect the comprehensive relationship between the process of business service to customer service and the satisfaction result in various aspects, thereby obtaining a target quality score with high accuracy.

[0099] Based on the technical solutions provided in the above embodiments, an optional embodiment is provided, in which the steps for determining the target evaluation weights are refined.

[0100] See Figure 5 The steps for determining the target evaluation weights include:

[0101] S510 retrieves historical service data and rating tags for different business services within a historical time period from the power service system.

[0102] The historical time period can be selected as needed, such as the past year. Of course, other time periods can also be selected, and there are no restrictions here.

[0103] Historical service data for business services can be understood as the service data generated by business services within a historical time period.

[0104] Among them, rating tags can be understood as tag data for manually rating business services.

[0105] S520 constructs a fitness function with evaluation weight as the independent variable and minimizing the scoring error, based on historical service data and rating labels.

[0106] The scoring error can be understood as the difference between the historical target quality score calculated using historical service data and the scoring label.

[0107] S530 uses a genetic algorithm to determine the evaluation weights of each objective based on the fitness function.

[0108] In the genetic algorithm, multiple individuals are set up, each corresponding to a combination of evaluation weights for various evaluation indicators. During the evolutionary process, selection, crossover, and mutation operations are performed on the individuals sequentially. The selection operation uses a tournament selection method, randomly selecting multiple individuals for fitness comparison and retaining the one with the highest fitness to enter the next generation. The crossover operation uses positional random crossover, exchanging some gene loci in the historical service feature data of an individual to generate a new individual. The mutation operation randomly assigns non-zero integer values ​​to the gene loci of an individual to maintain diversity. During the iteration process, the individual with the best fitness in the current population is retained until the maximum number of iterations or the convergence condition is reached. The final optimal individual output, corresponding to the combination of evaluation weights, serves as the target evaluation weight for each evaluation indicator.

[0109] In this embodiment, a fitness function is constructed with evaluation weight as the independent variable and minimizing the scoring error as the objective. A genetic algorithm is used to solve the fitness function, that is, the genetic algorithm is introduced to optimize in the weight search space, guiding the evaluation weight to dynamically find the optimal value, thereby obtaining the accurate target evaluation weight and improving the accuracy of the second quality score.

[0110] Based on the technical solutions provided in the above embodiments, an optional embodiment is provided, in which the fitness function construction step in S520 is refined.

[0111] See Figure 6The fitness function construction steps include:

[0112] S610, determine the first historical quality score of the business service based on historical service data and a preset service quality evaluation model.

[0113] That is, the historical service feature data corresponding to the historical service data of a business service is input into the service quality evaluation model to obtain the first historical quality score of the business service.

[0114] S620: Construct a fitness function based on the first historical quality score, rating labels, historical service data, and the initial values ​​of each evaluation weight.

[0115] For example, the fitness function is expressed as follows:

[0116]

[0117] In the formula, Here, M is the fitness function; M is the number of historical service data, i.e., the number of samples. The first historical quality score corresponding to the i-th historical service data; This represents the weight vector formed by the evaluation weights; T is the transpose symbol. For the i-th historical service data; Let be the rating label corresponding to the i-th historical service data.

[0118] The negative sign before the outermost parentheses in the above expression is used to transform the minimum error problem into a maximum fitness problem. The higher the fitness, the closer the historical target quality score under the weight vector is to the score label, and the more accurate the weight vector is.

[0119] In practical scenarios, the sum of the evaluation weights corresponding to each evaluation index is 1. To facilitate calculation and the computation and encoding operations of the genetic algorithm, a multi-parameter concatenated integer encoding method can be used to encode the weight vector before the genetic algorithm, resulting in an integer weight vector. Each weight in the integer weight vector is an integer, and this integer weight vector is used in the genetic algorithm. This encoding method not only satisfies the constraint of weight normalization but also avoids the stability problems caused by real-number precision calculations. After obtaining the optimal individual and thus the optimal integer weight vector using the genetic algorithm, the optimal integer weight vector is normalized to obtain the usable target evaluation weights. The normalization formula is:

[0120]

[0121] In the formula, A weight vector with real-valued precision. It is an integer weight vector, that is, a weight vector with integer precision.

[0122] The initial values ​​of each evaluation weight serve as the starting point for the genetic algorithm to dynamically optimize the weights.

[0123] In this embodiment, a suitable fitness function is constructed based on the first historical quality score, score labels, historical service data, and the initial values ​​of each evaluation weight. This fitness function is optimized by combining neural network prediction scores to improve the accuracy of the target evaluation index.

[0124] Based on the technical solutions provided in the above embodiments, an optional embodiment is provided. In this optional embodiment, a quality scoring method is provided, the method comprising:

[0125] S1, training the service quality evaluation model.

[0126] S2 retrieves historical service data and rating tags for different business services within a historical time period from the power service system.

[0127] S3. Input the historical service feature data corresponding to the historical service data into the service quality evaluation model to determine the first historical quality score of the power service system.

[0128] S4. Construct a fitness function based on the first historical quality score, rating labels, historical service data, and the initial values ​​of each evaluation weight.

[0129] S5 uses a genetic algorithm to determine the evaluation weights of each objective based on the fitness function.

[0130] S6: Obtain the service data to be evaluated corresponding to different business services of the power service system within a preset time period.

[0131] S7. For each business service, feature extraction is performed on the service data to be evaluated corresponding to the business service to obtain the corresponding service feature data.

[0132] S8 inputs the service feature data into the service quality evaluation model to obtain the first quality score of the business service.

[0133] S9. Multiply each feature value in the service feature data with the target evaluation weight of the corresponding evaluation indicator in each evaluation indicator to obtain the multiplication result corresponding to each feature value; sum up the multiplication results to obtain the second quality score.

[0134] S10, determine the target quality score for the business service based on the first quality score and the second quality score.

[0135] During the model training phase, the complex nonlinear relationship between service quality and customer satisfaction can be learned through a multilayer perceptron neural network, which can then output accurate score predictions.

[0136] Among them, the genetic algorithm was used to optimize the evaluation weight configuration of each evaluation indicator and dynamically adjust the contribution of each evaluation indicator under different business scenarios and service types, thereby effectively solving the evaluation distortion problem caused by fixed weighting coefficients in traditional scoring methods.

[0137] The target quality score is determined based on the first and second quality scores, combining the predictive power of neural networks with the global optimization characteristics of genetic algorithms. This allows for personalized weight adjustment and score output across different service types and customer groups, demonstrating strong adaptability and flexibility. Through this fusion mechanism, the customer service quality score not only possesses high predictive accuracy but also responds in real-time to changes in service processes and business needs. Furthermore, dynamically adjusting the weights in the weighting process of the first and second quality scores generates highly interpretable service quality evaluation results, facilitating further analysis and optimization by the customer service management system.

[0138] In the feature extraction process, structured data and unstructured text content can be comprehensively utilized. The TF-IDF method is used to perform deep semantic modeling of customer feedback and service communication records, and then fused with multi-dimensional structured features to construct a unified vector representation. This fusion approach not only improves the modeling capability of customer service satisfaction but also ensures the quantifiability and accuracy of rating results across multiple business scenarios.

[0139] Through actual deployment and testing, this embodiment demonstrates advantages such as high accuracy, strong generalization ability, and low latency in multiple power customer service scenarios. The overall solution outperforms existing traditional methods based on static weighting and empirical rules in terms of scoring accuracy, business adaptability, and system flexibility. Addressing the problems of single evaluation standards, poor dynamic adaptability, and excessive manual intervention in the current power industry customer service evaluation system, this embodiment constructs a complete intelligent evaluation process from data preprocessing, feature extraction and modeling, weight optimization to evaluation result output. This provides an important technical foundation for service resource allocation. In addition to its use in service resource allocation, the target quality score can also provide stable and efficient intelligent support for service quality monitoring, customer satisfaction improvement, and service strategy optimization. Furthermore, it provides an important technical foundation for the knowledge accumulation and evolution of intelligent customer service and service quality assessment systems.

[0140] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0141] Based on the same inventive concept, this application also provides a quality evaluation device for implementing the quality evaluation method described above. The solution provided by this device is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more quality evaluation device embodiments provided below can be found in the limitations of the quality evaluation method described above, and will not be repeated here.

[0142] In one exemplary embodiment, a quality assessment device is provided, see [link to relevant documentation]. Figure 7 It includes a data acquisition module 710, a feature extraction module 720, and a score determination module 730, wherein:

[0143] Data acquisition module 710 is used to acquire service data to be evaluated corresponding to different business services of the power service system within a preset time period; different business services are implemented using different business resources;

[0144] The feature extraction module 720 is used to extract features from the service data to be evaluated corresponding to each business service to obtain the corresponding service feature data.

[0145] The scoring determination module 730 is used to determine the target quality score of the business service based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service; wherein, the target quality score of each business service is used to schedule the service resources of the power service system.

[0146] In one embodiment, the apparatus further includes a weight determination module for determining the evaluation weights of each target, the weight determination module comprising:

[0147] The first acquisition unit is used to acquire historical service data and rating tags of different business services of the power service system within a historical time period;

[0148] The first building unit is used to construct a fitness function with evaluation weight as the independent variable and minimizing the scoring error as the objective, based on historical service data and rating labels.

[0149] The weight determination unit is used to determine the evaluation weight of each target based on the fitness function using a genetic algorithm.

[0150] In one embodiment, the first construction unit is specifically used to: determine a first historical quality score for the business service based on historical service data and a preset service quality evaluation model; and construct a fitness function based on the first historical quality score, score labels, historical service data, and the initial values ​​of each evaluation weight.

[0151] In one embodiment, the scoring determination module includes:

[0152] The model scoring unit is used to determine the first quality score of the business service based on the service feature data and the preset service quality evaluation model.

[0153] The indicator scoring unit is used to determine the second quality score of the business service based on the service feature data and the target evaluation weights corresponding to each evaluation indicator.

[0154] The target determination unit is used to determine the target quality score of the business service based on the first quality score and the second quality score.

[0155] In one embodiment, the indicator scoring unit is specifically used to: multiply each feature value in the service feature data by the target evaluation weight of the corresponding evaluation indicator in each evaluation indicator to obtain the multiplication result corresponding to each feature value; and accumulate the multiplication results to obtain the second quality score.

[0156] In one embodiment, the feature extraction module is specifically used to: standardize the structured data in the service data to be evaluated to obtain first feature data; encode the unstructured data in the service data to be evaluated to obtain second feature data; and concatenate and fuse the first feature data and the second feature data to obtain service feature data.

[0157] Each module in the aforementioned quality evaluation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0158] In one exemplary embodiment, a computer device is provided, the internal structure of which can be as shown in the figure. Figure 8As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements an intra-frame prediction mode determination method.

[0159] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0160] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the quality evaluation methods provided in the above embodiments.

[0161] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the quality evaluation methods provided in the above embodiments.

[0162] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the quality evaluation methods provided in the above embodiments.

[0163] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0164] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0165] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0166] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A quality evaluation method, characterized in that, include: Acquire service data to be evaluated for different business services within a preset time period from the power service system; different business services are implemented using different business resources; For each business service, feature extraction is performed on the service data to be evaluated corresponding to the business service to obtain the corresponding service feature data. Based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service, the target quality score of the business service is determined; wherein, the target quality score of each business service is used to schedule the service resources of the power service system.

2. The method according to claim 1, characterized in that, The evaluation weights for each of the aforementioned targets are determined in the following manner: Obtain historical service data and rating tags for different business services of the power service system within a historical time period; Based on the historical service data and the rating labels, a fitness function is constructed with the evaluation weight as the independent variable and the goal of minimizing the rating error. A genetic algorithm is used to determine the evaluation weights of each target based on the fitness function.

3. The method according to claim 2, characterized in that, The step of constructing a fitness function based on the historical service data and the rating labels, with evaluation weights as independent variables and minimizing rating error as the objective, includes: Based on the historical service data and the preset service quality evaluation model, the first historical quality score of the power service system is determined; The fitness function is constructed based on the first historical quality score, the score label, the historical service data, and the initial values ​​of each of the evaluation weights.

4. The method according to any one of claims 1 to 3, characterized in that, The step of determining the target quality score of the business service based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service includes: Based on the service feature data and the preset service quality evaluation model, a first quality score for the business service is determined; The second quality score of the business service is determined based on the service feature data and the target evaluation weights corresponding to each of the evaluation indicators. The target quality score of the business service is determined based on the first quality score and the second quality score.

5. The method according to claim 4, characterized in that, The step of determining the second quality score of the business service based on the service feature data and the target evaluation weights corresponding to each of the evaluation indicators includes: Each feature value in the service feature data is multiplied by the target evaluation weight of the corresponding evaluation indicator in each of the evaluation indicators to obtain the multiplication result corresponding to each feature value; The results of each multiplication are summed to obtain the second quality score.

6. The method according to claim 1, characterized in that, The step of extracting features from the service data to be evaluated to obtain service feature data includes: The structured data in the service data to be evaluated is standardized to obtain the first feature data; The unstructured data in the service data to be evaluated is encoded to obtain the second feature data; The first feature data and the second feature data are concatenated and merged to obtain the service feature data.

7. A quality evaluation device, characterized in that, include: The data acquisition module is used to acquire service data to be evaluated for different business services in the power service system within a preset time period; different business services are implemented using different business resources. The feature extraction module is used to extract features from the service data to be evaluated corresponding to each business service to obtain the corresponding service feature data. The scoring determination module is used to determine the target quality score of the business service based on the service feature data and the target evaluation weights corresponding to multiple evaluation indicators under the business service; wherein, the target quality score of each business service is used to schedule the service resources of the power service system.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images