Service providing user evaluation method, device, equipment, medium and program product
Through the evaluation model of feature extraction and random forest algorithm training, and the decision tree is determined in combination with genetic algorithms, the problem of inaccurate service quality assessment in the existing technology is solved, and a more accurate and objective service quality assessment is achieved.
Patent Information
- Application Number
- CN202411159196.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-08
AI Technical Summary
In the prior art, the service quality evaluation results of the service provider user are not accurate enough, and are greatly affected by subjective factors, resulting in inaccurate evaluation results.
By obtaining the attribute data of the target service based on the preset evaluation dimension, performing feature extraction processing, inputting it into the evaluation model, using a random forest algorithm to train the evaluation model, combining the genetic algorithm to determine the decision tree, generate evaluation results, and avoiding the influence of subjective factors.
It improves the accuracy of the service providing user evaluation results, reduces the dimension of attribute data, and retains key attribute information, providing a more objective service quality assessment.
Smart Images

Figure CN118916748A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a method, device, equipment, medium and program product for evaluating service provider users. Background Art
[0002] Service providers are common in many fields. However, it is necessary to evaluate the services provided by service providers to determine the service quality of each service provider.
[0003] In the prior art, evaluation dimensions of services provided by service providers can be pre-configured, and service data corresponding to each evaluation dimension can be obtained based on the configured evaluation dimensions, and the obtained service data can be processed based on pre-set rules to determine the service quality of each service provider. However, the pre-set rules are set based on the experience of expert users, are affected by subjective factors, and are not accurate enough, resulting in inaccurate evaluation results when the service quality of service providers is evaluated based on the pre-set rules. Summary of the invention
[0004] Based on this, it is necessary to provide a service provider user evaluation method, device, equipment, medium and program product that can improve the accuracy of evaluation results in response to the above technical problems.
[0005] In a first aspect, the present application provides a service providing user evaluation method, comprising:
[0006] Acquire multiple first attribute data of the target service provider user based on preset evaluation dimensions, where each first attribute data corresponds to a different evaluation dimension;
[0007] Performing feature extraction processing on each first attribute data to obtain a plurality of target attribute data;
[0008] Multiple target attribute data are input into the evaluation model, and the evaluation result of the target service provider is determined according to the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider and the dimension score corresponding to each target attribute data.
[0009] In one of the embodiments, the method further includes: determining a service quality level of the target service provider according to an evaluation result of the target service provider.
[0010] In one of the embodiments, determining the service quality level of the target service providing user based on the evaluation results of the target service providing user includes: obtaining the evaluation results of multiple service providing users, determining the ranking of the target service providing user among the multiple service providing users based on the evaluation results of each service providing user; and determining the service quality level of the target service providing user based on the ranking of the target service providing user.
[0011] In one of the embodiments, the evaluation dimensions corresponding to the target attribute data include: the number of registered users attracted, the transaction amount, the number of transaction customers, the number of orders, the number of times merchants log into the mall, the order confirmation speed, the order shipment speed, the order delivery speed, the order receipt speed, the number of products removed from the shelves due to inspections, and the number of effective penalty orders for service provider users.
[0012] In one embodiment, the training process of the evaluation model includes: obtaining multiple first training attribute data of the training service provider based on a preset evaluation dimension, each first training attribute data corresponds to a different evaluation dimension; performing feature extraction processing on each first training attribute data to obtain multiple target training attribute data; obtaining training evaluation results of the training service provider; using the multiple target training attribute data and the training evaluation results as training samples to train the initial evaluation model to obtain the evaluation model.
[0013] In one of the embodiments, the evaluation model is determined based on a random forest algorithm, and multiple target training attribute data and training evaluation results are used as training samples to train the initial evaluation model to obtain the evaluation model, including: sampling multiple sub-training sample sets from the training samples with replacement, wherein the number of sub-training sample sets is determined based on a genetic algorithm; determining a decision tree based on each sub-training sample set; determining a prediction result of the initial evaluation model based on the output result of each decision tree, and training the initial evaluation model based on the prediction result and the training evaluation result.
[0014] In a second aspect, the present application also provides a service providing user evaluation device, including:
[0015] An acquisition module, used to acquire a plurality of first attribute data of a target service providing user based on a preset evaluation dimension, wherein each first attribute data corresponds to a different evaluation dimension;
[0016] A processing module, used for performing feature extraction processing on each first attribute data to obtain a plurality of target attribute data;
[0017] The evaluation module is used to input multiple target attribute data into the evaluation model, and determine the evaluation result of the target service provider user based on the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimensional score corresponding to each target attribute data.
[0018] In one of the embodiments, the evaluation module is further used to determine the service quality level of the target service provider according to the evaluation result of the target service provider.
[0019] In one of the embodiments, the evaluation module is further used to obtain evaluation results of multiple service providing users, determine the ranking of the target service providing user among the multiple service providing users based on the evaluation results of each service providing user; and determine the service quality level of the target service providing user based on the ranking of the target service providing user.
[0020] In one of the embodiments, the evaluation dimensions corresponding to the target attribute data include: the number of registered users attracted, the transaction amount, the number of transaction customers, the number of orders, the number of times merchants log into the mall, the order confirmation speed, the order shipment speed, the order delivery speed, the order receipt speed, the number of products removed from the shelves due to inspections, and the number of effective penalty orders for service provider users.
[0021] In one of the embodiments, the service provider user evaluation device also includes a model training module, which is used to obtain multiple first training attribute data of the training service provider based on preset evaluation dimensions, and each first training attribute data corresponds to a different evaluation dimension; perform feature extraction processing on each first training attribute data to obtain multiple target training attribute data; obtain training evaluation results of the training service provider; use the multiple target training attribute data and the training evaluation results as training samples to train the initial evaluation model to obtain an evaluation model.
[0022] In one of the embodiments, the evaluation model is determined based on a random forest algorithm, and the model training module is specifically used to sample multiple sub-training sample sets from the training samples with replacement, wherein the number of sub-training sample sets is determined based on a genetic algorithm; a decision tree is determined based on each sub-training sample set; a prediction result of the initial evaluation model is determined based on the output result of each decision tree, and the initial model is trained based on the prediction result and the training evaluation result.
[0023] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, any of the methods described in the first aspect is implemented.
[0024] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any of the methods described in the first aspect above.
[0025] In a fifth aspect, the present application further provides a computer program product, including a computer program, which, when executed by a processor, implements any of the methods described in the first aspect above.
[0026] The above-mentioned service provider user evaluation method, device, equipment, medium and program product first obtain multiple first attribute data of the target service provider user based on preset evaluation dimensions, and each first attribute data corresponds to a different evaluation dimension. Then, feature extraction processing is performed on each first attribute data to obtain multiple target attribute data. Finally, the multiple target attribute data are input into the evaluation model, and the evaluation result of the target service provider user is determined according to the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimensional score corresponding to each target attribute data. In this way, by performing feature extraction processing on the acquired first attribute data, the loss of the information contained therein can be minimized while reducing the number of first attribute data, that is, the obtained target attribute data reduces the dimension of the data set, while retaining as much attribute information of the target service provider as possible, and taking the target attribute data as input, the evaluation result of the target service provider user can be accurately determined through the pre-trained evaluation model to avoid the influence of subjective factors on the evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0028] Figure 1 A diagram of an application environment for a method for providing user evaluation of a service in an embodiment;
[0029] Figure 2 A flowchart of a method for providing user evaluation of a service in one embodiment;
[0030] Figure 3 A flow chart of the steps of determining the service quality level of a target service provider in one embodiment;
[0031] Figure 4 A schematic diagram of a training step flow of an evaluation model in one embodiment;
[0032] Figure 5 A flowchart of the training steps of an evaluation model in another embodiment;
[0033] Figure 6 A flowchart of a method for providing user evaluation of a service in another embodiment;
[0034] Figure 7 A structural block diagram of a service providing user evaluation device in one embodiment;
[0035] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0036] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0037] Service providers are common in many fields. However, it is necessary to evaluate the services provided by service providers to determine the service quality of each service provider.
[0038] In the prior art, evaluation dimensions of services provided by service providers can be pre-configured, and service data corresponding to each evaluation dimension can be obtained based on the configured evaluation dimensions, and the obtained service data can be processed based on pre-set rules to determine the service quality of each service provider. However, the pre-set rules are set based on the experience of expert users, are affected by subjective factors, and are not accurate enough, resulting in inaccurate evaluation results when the service quality of service providers is evaluated based on the pre-set rules.
[0039] In view of this, the present application provides a service provider user evaluation method that can improve the accuracy of the evaluation result. The service provider user evaluation method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The service provider user evaluation method provided in the embodiment of the present application can be applied to the server 104, and can also be implemented through the interaction between the terminal 102 and the server 104. For example, the server 104 obtains multiple first attribute data of the target service provider user based on a preset evaluation dimension, and each first attribute data corresponds to a different evaluation dimension, and performs feature extraction processing on each first attribute data to obtain multiple target attribute data, and then inputs the multiple target attribute data into the evaluation model, and determines the evaluation result of the target service provider user according to the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimensional score corresponding to each target attribute data. Finally, the evaluation result of the target service provider user can be sent to the terminal 102 to display the evaluation result of the target service provider user through the terminal 102. The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 may be an independent physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.
[0040] In an exemplary embodiment, Figure 2 As shown, a service provision user evaluation method is provided, and the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 201 to 203. Among them:
[0041] S201, the server obtains a plurality of first attribute data of a target service providing user based on a preset evaluation dimension.
[0042] Among them, the evaluation dimensions corresponding to each first attribute data are different.
[0043] Optionally, the target service provider may be a service provider whose service quality is to be evaluated, and the service provider may have already provided services or may be a candidate service provider to be provided services.
[0044] Optionally, the preset evaluation dimensions are configured in advance in a database or configuration file, and when business personnel need to adjust the evaluation dimensions, they can do so by modifying the configuration in the database or configuration file, so that the evaluation dimensions can be flexibly adjusted when the management system changes.
[0045] In one possible implementation, the first attribute data of the target service provider user may be obtained from a database in an e-commerce system or an external system. When the first attribute data of the target service provider user is obtained from the database, the acquisition method is simple.
[0046] In another possible implementation, the first attribute data of the target service provider user may be determined by a behavioral analysis of the target service user. For example, the behavioral data of the target service user in the system may be obtained through a crawler tool, and the behavioral data may be input into a preset analysis model to obtain multiple attribute data of the target service provider user, and the first attribute data may be filtered out from the multiple attribute data according to a preset evaluation dimension.
[0047] S202: The server performs feature extraction processing on each first attribute data to obtain a plurality of target attribute data.
[0048] Optionally, before performing feature extraction processing on each first attribute data, the first attribute data also needs to be cleaned, deduplicated and standardized to ensure the accuracy and consistency of the data, wherein data cleaning of the first attribute data may include processing missing values and outliers, and standardization of the first attribute data may convert data of different dimensions to the same scale to eliminate the dimensional influence between different features. The first attribute data may be standardized by normalization or standard deviation standardization, which is not limited in this embodiment of the present application.
[0049] In one possible implementation, feature extraction processing is performed on each first attribute data, and the information gain value corresponding to each first attribute data can be determined through an information gain function, and a portion of features with larger information gain values can be retained as target attribute data. For example, the information gain values corresponding to each first attribute data are sorted in order from large to small, and the top multiple first attribute data are used as target attribute data.
[0050] In another possible implementation, feature extraction processing for each first attribute data may be implemented through principal component analysis, which reduces the dimension of each first attribute data to reduce the number of features and retain important information.
[0051] Optionally, the evaluation dimensions corresponding to the target attribute data may include: the number of registered users attracted, the transaction amount, the number of transaction customers, the number of orders, the number of times merchants log into the mall, the order confirmation speed, the order shipment speed, the order delivery speed, the order receipt speed, the number of products removed from the shelves due to inspections, and the number of effective penalty orders for service provider users.
[0052] S203, the server inputs a plurality of target attribute data into the evaluation model, and determines the evaluation result of the target service providing user according to the output of the evaluation model.
[0053] The evaluation results are used to represent the comprehensive scores of the target service provider users and the dimensional scores corresponding to each target attribute data.
[0054] Optionally, the evaluation model can be a model pre-trained through machine learning that can determine the comprehensive score of the target service provider and the dimension scores corresponding to each target attribute data based on the target attribute data of the target service provider. The input of the model is the target attribute data of the target service provider, and the evaluation result of the target service provider is determined based on the output of the model.
[0055] In one possible implementation, the evaluation model can be based on a logistic regression algorithm, which is very fast to train and has very good interpretability.
[0056] In another possible implementation, the evaluation model can be implemented based on a decision tree algorithm, which is a non-parametric supervised learning method that can summarize decision rules from a series of data with attributes and labels.
[0057] In another possible implementation, the evaluation model can be implemented based on the random forest algorithm, which improves accuracy and reduces overfitting by combining multiple decision trees.
[0058] In the above-mentioned service provider user evaluation method, first, multiple first attribute data of the target service provider user are obtained based on preset evaluation dimensions, and each first attribute data corresponds to a different evaluation dimension. Then, feature extraction processing is performed on each first attribute data to obtain multiple target attribute data. Finally, the multiple target attribute data are input into the evaluation model, and the evaluation result of the target service provider user is determined according to the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimensional score corresponding to each target attribute data. In this way, by performing feature extraction processing on the acquired first attribute data, the loss of the information contained therein can be minimized while reducing the number of first attribute data. That is, the obtained target attribute data reduces the dimension of the data set, while retaining as much attribute information of the target service provider as possible. The target attribute data is used as input, and the evaluation result of the target service provider user can be accurately determined through a pre-trained evaluation model to avoid the influence of subjective factors on the evaluation result.
[0059] In an exemplary embodiment, optionally, the service providing user evaluation method further includes: determining a service quality level of the target service providing user according to an evaluation result of the target service providing user.
[0060] Optionally, after determining the evaluation result of the target service provider, the service quality level of the target service provider can be determined according to the comprehensive score of the target service provider and the dimension scores corresponding to each target attribute data.
[0061] In one possible implementation, the scoring intervals corresponding to the preset service quality levels can be obtained to determine in which scoring interval the comprehensive score of the target service provider falls, and then the service quality level of the target service provider is determined. Alternatively, the scoring interval of the dimension score corresponding to each target attribute data is determined, and the level of each dimension of the target service provider is determined. Then, the service quality level of the target service provider is determined based on the level of each dimension.
[0062] In another possible implementation, the evaluation results of multiple service providing users may be ranked to determine the service quality level of the target service providing user according to the ranking of the target service providing user in the evaluation results of the multiple service providing users.
[0063] Optionally, the service quality level of the target service provider can be determined based on the comprehensive score of the target service provider, or based on the dimension scores corresponding to each target attribute data and the dimension weights corresponding to each target attribute data, or the service quality level of the target service provider can be determined based on both methods, that is, the first level of the target service provider is determined by the comprehensive score, and the second level of the target service provider is determined by the dimension scores and dimension weights corresponding to each target attribute data, and then the service quality level of the target service provider is determined based on the first level and the second level, which can be achieved by weighted summation or average value, or the lowest level between the first level and the second level is selected as the service quality level of the target service provider. The embodiment of the present application does not limit the method for determining the service quality level of the target service provider based on the first level and the second level.
[0064] Exemplarily, the comprehensive score of service provider user A is 85 points, the score for the number of registered users attracted is 90 points, the score for the transaction amount is 80 points, the score for the number of transaction customers is 85 points, the score for the number of orders is 90 points, the score for the number of merchant logins to the mall is 80 points, the score for the order confirmation speed is 75 points, the score for the order shipment speed is 85 points, the score for the order successful delivery speed is 80 points, the score for the order receipt speed is 85 points, the score for the number of goods removed from the shelves due to inspections is 90 points, the score for the number of effective penalty orders for service provider users is 70 points, and the service quality level is five stars; the comprehensive score of service provider user B is 78 points, the score for the number of registered users attracted is 75 points, the score for the transaction amount is 80 points, the score for the number of transaction customers is 70 points, the score for the number of orders is 75 points, the score for the number of merchant logins to the mall is 70 points, the score for the order confirmation speed is 80 points, the score for the order shipment speed is 79 points, the score for the order successful delivery speed is 75 points, the score for the order receipt speed is 78 points, the score for the number of goods removed from the shelves due to inspections is 85 points, the score for the number of effective penalty orders for suppliers is 65 points, and the service quality level is four stars.
[0065] In an exemplary embodiment, Figure 3 As shown, based on the above embodiment, optionally, according to the evaluation result of the target service provider, the service quality level of the target service provider is determined, including the following steps 301 to 302. Among them:
[0066] S301, the server obtains evaluation results of multiple service providing users, and determines the ranking of the target service providing user among the multiple service providing users according to the evaluation results of each service providing user.
[0067] Optionally, after determining the evaluation result of the target service providing user, the evaluation results of other service providing users may be obtained, wherein the evaluation results of other service providing users are also determined based on the method in the above embodiment and stored in the database.
[0068] Optionally, after obtaining the evaluation results of multiple service providers including the target service provider, the comprehensive scores of the multiple service providers can be sorted to determine the ranking of the target service provider; or the dimension scores corresponding to the target attribute data of the multiple service providers and the dimension weights corresponding to the target attribute data can be obtained, and a score value can be determined by weighted summation, and the score values can be sorted to determine the ranking of the target service provider; or multiple rankings of the target service provider can be determined by sorting the dimension scores corresponding to the target attribute data, and the lowest ranking among the multiple rankings can be used as the ranking of the target service provider among the multiple service providers, or the highest ranking among the multiple rankings can be used as the ranking of the target service provider among the multiple service providers, or the ranking of the dimension score with the largest dimension weight corresponding to the target attribute data can be used as the ranking of the target service user among the multiple service providers. The embodiment of the present application does not limit the method for determining the ranking of the target service provider among the multiple service providers.
[0069] Optionally, the sorting may be in positive order, that is, the higher the score, the higher the ranking.
[0070] S302: The server determines the service quality level of the target service provider according to the ranking of the target service provider.
[0071] Optionally, the service quality level is determined according to a preset service quality level rule and the ranking of the target service provider. For example, if the target service provider is ranked in the top 10% of multiple service providers, the service quality level of the target service provider is five stars, and the service quality of the target service provider is optimal. If the target service provider is ranked in the top 10%-30% of multiple service providers, the service quality level of the target service provider is four stars; if the target service provider is ranked in the top 30%-50% of multiple service providers, the service quality level of the target service provider is three stars; if the target service provider is ranked in the top 50%-80% of multiple service providers, the service quality level of the target service provider is two stars; if the target service provider is ranked in the top 80%-100% of multiple service providers, the service quality level of the target service provider is one star.
[0072] Optionally, the evaluation results and service quality levels of the target service provider users may be recorded in a database, and displayed in the form of charts, reports, etc., providing multi-dimensional analysis and comparison functions.
[0073] The above-mentioned evaluation results of multiple service providers are obtained, the ranking of the target service provider among the multiple service providers is determined according to the evaluation results of each service provider, and the service quality level of the target service provider is determined according to the ranking of the target service provider. The service quality level of the target service provider can be clearly and accurately determined according to the evaluation results. The enterprise can clearly understand the advantages and disadvantages of each service provider based on the comprehensive score and the scores of each dimension in the evaluation results as well as the service quality level, accurately determine the service quality of each service provider, and better manage and select service providers.
[0074] In an exemplary embodiment, Figure 4 As shown, optionally, the training process of the evaluation model includes the following steps 401 to 404. Among them:
[0075] S401: The server obtains a plurality of first training attribute data of a user providing a training service based on a preset evaluation dimension.
[0076] Among them, the evaluation dimensions corresponding to each first training attribute data are different.
[0077] Optionally, the first training attribute data refers to original data used as input data during the evaluation model training process and is used to train the model.
[0078] Optionally, when obtaining the first training attribute data, the data of the corresponding evaluation dimension provided by the training service to the user may be obtained by querying the training service or by crawling and analyzing and screening the collected data.
[0079] S402: The server performs feature extraction processing on each first training attribute data to obtain a plurality of target training attribute data.
[0080] Optionally, before performing feature extraction processing on each first training attribute data, it is also necessary to perform data cleaning, deduplication and standardization processing on the first attribute data.
[0081] Optionally, the manner of performing feature extraction processing on the first training attribute data is consistent with the manner of performing feature extraction processing on the first attribute data in step 202 in the above embodiment, and will not be described in detail in the embodiment of the present application.
[0082] S403: The server obtains the training evaluation result provided by the training service provider to the user.
[0083] Optionally, the training evaluation result of the training service provider may be obtained from the database, and the training evaluation result is obtained by evaluating each first training attribute data using an objective evaluation method.
[0084] In a possible implementation, each first training attribute data may be evaluated by using the Topsis comprehensive evaluation method to obtain a training evaluation result.
[0085] In another possible implementation manner, each first training attribute data may be evaluated by a grey correlation analysis method to obtain a training evaluation result.
[0086] S404: The server uses the multiple target training attribute data and the training evaluation results as training samples to train the initial evaluation model to obtain an evaluation model.
[0087] Optionally, an initial evaluation model may be established based on a logistic regression algorithm, a logic tree algorithm, or a random forest algorithm.
[0088] In one possible implementation, 80% of the training sample data is used to train the model, and 20% of the training sample data is used to test the correctness of the model. If the evaluation model obtained does not achieve the expected effect, the parameters are adjusted for continuous optimization to eventually obtain a model with higher accuracy.
[0089] In another possible implementation method, part of the training sample data is used as the initial training set, and the initial training set is used to train the evaluation model. If the obtained evaluation model does not achieve the expected effect, some samples with errors in the evaluation model are added to the initial training set, or some samples with correct evaluation models are removed from the initial training set, and the evaluation model is retrained using the updated initial training set until the evaluation model achieves the expected effect or the training end condition is met.
[0090] During the training process of the above-mentioned evaluation model, multiple first training attribute data of the training service provider are obtained based on the preset evaluation dimension, and the evaluation dimensions corresponding to each first training attribute data are different; feature extraction processing is performed on each first training attribute data to obtain multiple target training attribute data; the training evaluation results of the training service provider are obtained; the multiple target training attribute data and the training evaluation results are used as training samples to train the initial evaluation model to obtain the evaluation model. In this way, the dimension of the training sample is reduced while retaining as much attribute information as possible, the initial model is trained, and the accuracy of the evaluation model is improved. Therefore, after the target attribute data of the target service provider is input into the evaluation model, the evaluation result of the target service provider can be accurately determined according to the output of the model.
[0091] In an exemplary embodiment, Figure 5 As shown, optionally, the evaluation model is determined based on a random forest algorithm, multiple target training attribute data and training evaluation results are used as training samples, and the initial evaluation model is trained to obtain the evaluation model, including the following steps 501 to 503. Among them:
[0092] S501 : The server samples multiple sub-training sample sets from the training samples with replacement, wherein the number of the sub-training sample sets is determined based on a genetic algorithm.
[0093] Optionally, the random forest combines multiple decision trees. Determining the evaluation model based on the random forest algorithm may include two steps. The first part is random sampling, and the second part is the construction of a decision tree. A decision tree is generated based on a sub-training sample set. If m sub-training sample sets are sampled, m decision trees are constructed based on the m sub-training sample sets.
[0094] Optionally, in the random sampling part, each decision tree needs to sample a portion of the training samples. Bootstrap random sampling can be used to sample with replacement from the training samples. Bootstrap random sampling can effectively reduce the risk of overfitting of the decision tree.
[0095] Optionally, in random forests, the number of decision trees will affect the results. When the number of decision trees is small, random forests are prone to underfitting, and the evaluation model performs poorly on both training data and test data. When the number of decision trees is large, random forests are prone to overfitting, that is, the evaluation model performs well on training data but poorly on test data. Therefore, the selection of the number of decision trees is crucial.
[0096] Optionally, the number of decision trees can be determined based on a genetic algorithm, that is, the number of sub-training sample sets can be determined. The genetic algorithm can initialize the population and calculate the fitness of each individual, where each individual can represent the number of decision trees respectively, and then select some individuals as parents according to the fitness to generate the next generation of individuals. Two individuals can be selected from the parent generation, new individuals are generated through gene crossover operations, and gene mutation operations are performed on the newly generated individuals. The original individuals are then replaced with the generated new individuals to form a new population. When the number of iterations is reached or the pre-set optimal conditions are met, the current optimal solution is output as the final number of decision trees, where the accuracy of the evaluation model can be used as a fitness function to calculate the fitness of each individual.
[0097] S502: The server determines a decision tree based on each sub-training sample set.
[0098] Optionally, a portion of features can be selected from the sub-training sample set, and then these features can be used to construct the tree result, which can increase the randomness of the decision tree, reduce the correlation between decision trees, and improve the stability of the decision tree.
[0099] Optionally, the decision tree is constructed according to the CART decision tree algorithm, and the sub-training sample set is segmented by finding the optimal feature in the feature set to construct a decision tree.
[0100] S503, the server determines the prediction result of the initial evaluation model according to the output results of each decision tree, and trains the initial model according to the prediction result and the training evaluation result.
[0101] Optionally, the output results of each decision tree can be integrated in the form of an average value to obtain a final prediction result, and the prediction result can be compared with the training evaluation result to adjust the parameters in the initial evaluation model. For example, the parameters in the initial evaluation model can include the number of decision trees, the number of features considered when splitting nodes of the decision tree, and the maximum depth of the decision tree.
[0102] The above training samples include multiple sub-training sample sets sampled with replacement, wherein the number of sub-training sample sets is determined based on a genetic algorithm; a decision tree is determined based on each sub-training sample set; a prediction result of an initial evaluation model is determined based on the output result of each decision tree, the initial model is trained based on the prediction result and the training evaluation result, the optimal number of decision trees is determined through a genetic algorithm, and the initial training model is trained based on the training evaluation result to obtain an evaluation model with high evaluation accuracy, so that accurate evaluation results can be obtained when the service quality of users provided by the target service is evaluated through the evaluation model.
[0103] As an optional implementation, Figure 6 As shown, the service provider user evaluation method provided in the embodiment of the present application may include the following specific steps:
[0104] S601: The server obtains a plurality of first training attribute data of a user provided with a training service based on a preset evaluation dimension.
[0105] Among them, the evaluation dimensions corresponding to each first training attribute data are different.
[0106] S602: The server performs feature extraction processing on each first training attribute data to obtain a plurality of target training attribute data.
[0107] S603: The server obtains the training evaluation result provided by the training service to the user.
[0108] S604: The evaluation model is determined based on the random forest algorithm, and the server samples multiple sub-training sample sets from the training samples with replacement.
[0109] The number of sub-training sample sets is determined based on a genetic algorithm.
[0110] S605: The server determines a decision tree based on each sub-training sample set.
[0111] S606, the server determines the prediction result of the initial evaluation model according to the output results of each decision tree, and trains the initial model according to the prediction result and the training evaluation result to obtain the evaluation model.
[0112] S607: The server obtains a plurality of first attribute data of the target service providing user based on a preset evaluation dimension.
[0113] Among them, the evaluation dimensions corresponding to each first attribute data are different.
[0114] S608: The server performs feature extraction processing on each first attribute data to obtain multiple target attribute data.
[0115] S609, the server inputs the multiple target attribute data into the evaluation model, and determines the evaluation result of the target service provider according to the output of the evaluation model.
[0116] The evaluation results are used to represent the comprehensive scores of the target service provider users and the dimensional scores corresponding to each target attribute data.
[0117] S610: The server obtains evaluation results of multiple service providing users, and determines a ranking of the target service providing user among the multiple service providing users according to the evaluation results of each service providing user.
[0118] S611, the server determines the service quality level of the target service provider user according to the ranking of the target service provider user.
[0119] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0120] Based on the same inventive concept, the embodiment of the present application also provides a service provider user evaluation device for implementing the service provider user evaluation method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the one or more service provider user evaluation device embodiments provided below can refer to the limitations of the service provider user evaluation method above, and will not be repeated here.
[0121] In an exemplary embodiment, Figure 7 As shown, a service provider user evaluation device 700 is provided, comprising: an acquisition module 701, a processing module 702 and an evaluation module 703, wherein:
[0122] An acquisition module 701 is used to acquire a plurality of first attribute data of a target service providing user based on a preset evaluation dimension, wherein each of the first attribute data corresponds to a different evaluation dimension;
[0123] A processing module 702 is used to perform feature extraction processing on each of the first attribute data to obtain multiple target attribute data;
[0124] Evaluation module 703 is used to input the multiple target attribute data into the evaluation model, and determine the evaluation result of the target service provider user according to the output of the evaluation model, and the evaluation result is used to represent the comprehensive score of the target service provider user and the dimension score corresponding to each target attribute data.
[0125] In an exemplary embodiment, the evaluation module 703 is further configured to determine the service quality level of the target service provider according to the evaluation result of the target service provider.
[0126] In an exemplary embodiment, the evaluation module 703 is also used to obtain evaluation results of multiple service providing users, determine the ranking of the target service providing user among the multiple service providing users based on the evaluation results of each service providing user; and determine the service quality level of the target service providing user based on the ranking of the target service providing user.
[0127] In an exemplary embodiment, the evaluation dimensions corresponding to the target attribute data include: attracting the number of registered users, transaction amount, number of transaction customers, number of orders, number of times merchants log into the mall, order confirmation speed, order shipment speed, order delivery speed, order receipt speed, number of products removed from shelves due to inspections, and number of effective penalty orders for service provider users.
[0128] In an exemplary embodiment, the service provider user evaluation device 700 also includes a model training module, which is used to obtain multiple first training attribute data of the training service provider based on preset evaluation dimensions, and each first training attribute data corresponds to a different evaluation dimension; perform feature extraction processing on each first training attribute data to obtain multiple target training attribute data; obtain training evaluation results of the training service provider; use the multiple target training attribute data and the training evaluation results as training samples to train the initial evaluation model to obtain an evaluation model.
[0129] In an exemplary embodiment, the evaluation model is determined based on a random forest algorithm, and the model training module is specifically used to sample multiple sub-training sample sets from the training samples with replacement, wherein the number of sub-training sample sets is determined based on a genetic algorithm; a decision tree is determined based on each sub-training sample set; a prediction result of the initial evaluation model is determined based on the output result of each decision tree, and the initial model is trained based on the prediction result and the training evaluation result.
[0130] Each module in the above service providing user evaluation device can be implemented in whole or in part by software, hardware and their combination. Each module can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.
[0131] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store attribute data and evaluation results of each service provider. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a service provider user evaluation method is implemented.
[0132] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0133] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps described in any of the above method embodiments when executing the computer program.
[0134] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps described in any of the above method embodiments are implemented.
[0135] In an exemplary embodiment, a computer program product is provided, including a computer program, which implements the steps described in any of the above method embodiments when executed by a processor.
[0136] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present 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. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.
[0137] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, 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.
[0138] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
Claims
1. A service providing user evaluation method, characterized in that: The method comprises: Acquire multiple first attribute data of the target service provider user based on preset evaluation dimensions, where each of the first attribute data corresponds to a different evaluation dimension; Performing feature extraction processing on each of the first attribute data to obtain a plurality of target attribute data; The multiple target attribute data are input into the evaluation model, and the evaluation result of the target service provider user is determined according to the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimension score corresponding to each target attribute data.
2. The method according to claim 1, characterized in that: The method further comprises: The service quality level of the target service provider is determined according to the evaluation result of the target service provider.
3. The method according to claim 2, characterized in that Determining the service quality level of the target service provider user according to the evaluation result of the target service provider user includes: Obtaining evaluation results of a plurality of service providing users, and determining a ranking of a target service providing user among the plurality of service providing users according to the evaluation results of each of the service providing users; The service quality level of the target service provider user is determined according to the ranking of the target service provider user.
4. The method according to claim 1, characterized in that: The evaluation dimensions corresponding to the target attribute data include: the number of registered users attracted, transaction amount, number of transaction customers, number of orders, number of times merchants log into the mall, order confirmation speed, order shipment speed, order delivery speed, order receipt speed, number of products removed from shelves due to inspections, and number of effective penalty orders for service provider users.
5. The method according to claim 1, characterized in that The training process of the evaluation model includes: Acquire a plurality of first training attribute data of the training service provider based on a preset evaluation dimension, wherein each of the first training attribute data corresponds to a different evaluation dimension; Performing feature extraction processing on each of the first training attribute data to obtain a plurality of target training attribute data; Obtaining training evaluation results of users provided by the training service; The multiple target training attribute data and the training evaluation results are used as training samples to train the initial evaluation model to obtain the evaluation model.
6. The method according to claim 5, characterized in that The evaluation model is determined based on a random forest algorithm, and the multiple target training attribute data and the training evaluation results are used as training samples to train the initial evaluation model to obtain the evaluation model, including: Sampling a plurality of sub-training sample sets from the training sample with replacement, wherein the number of the sub-training sample sets is determined based on a genetic algorithm; Determine a decision tree based on each of the sub-training sample sets; The prediction result of the initial evaluation model is determined according to the output results of each decision tree, and the initial evaluation model is trained according to the prediction result and the training evaluation result.
7. A service providing user evaluation device, characterized in that: The device comprises: An acquisition module, used for acquiring a plurality of first attribute data of a target service providing user based on a preset evaluation dimension, wherein each of the first attribute data corresponds to a different evaluation dimension; A processing module, used for performing feature extraction processing on each of the first attribute data to obtain a plurality of target attribute data; An evaluation module is used to input the multiple target attribute data into an evaluation model, and determine the evaluation result of the target service provider user based on the output of the evaluation model. The evaluation result is used to represent the comprehensive score of the target service provider user and the dimensional score corresponding to each target attribute data.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
