Team service quality control method, device, equipment, medium and program product

Abstract

The invention relates to a team service quality control method and device, equipment, a medium and a program product. The method comprises the steps of determining a target service risk feature covered in current service data according to the current service data of a target service in a current sampling time period and each candidate service risk feature, determining a risk service team from each candidate service team according to the target service risk feature and the current service data, and sending the risk service team to the target service. And determining a target management and control mode of the risk service team according to the service data corresponding to the risk service team in the current service data and each standard service risk feature, and managing and controlling the service quality of the risk service team based on the target management and control mode. By adopting the method, the comprehensiveness of business quality control can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a team business quality control method, device, equipment, medium and program product. Background Art

[0002] With the continuous development of the financial field, the business volume of financial-related businesses has also increased. Therefore, in order to ensure the service quality of the financial business team, business quality control methods have emerged. Existing business quality control methods generally directly use personnel positioning models to locate abnormal business personnel based on business data.

[0003] However, due to the characteristics of team business data such as small data volume, large data fluctuations, and unbalanced data distribution, the existing business quality control methods can only perform business quality control on individual personnel, and cannot achieve overall business instruction control of the team, reducing the comprehensiveness of business quality control. Summary of the invention

[0004] Based on this, it is necessary to provide a team business quality control method, device, equipment, medium and program product that can improve the comprehensiveness of business quality control in response to the above technical problems.

[0005] In a first aspect, the present application provides a team business quality control method, including:

[0006] Determine the target business risk characteristics covered by the current business data according to the current business data of the target business and the risk characteristics of each candidate business within the current sampling period;

[0007] Determine the risk business team from the candidate business teams based on the target business risk characteristics and current business data;

[0008] Determine the target management and control method of the risk business team based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0009] Based on the target management method, the business quality of the risk business team is controlled.

[0010] In one embodiment, according to the current business data of the target business in the current sampling period and each candidate business risk feature, determining the target business risk feature covered by the current business data includes:

[0011] Feature extraction is performed on the historical business data of the target business to obtain the risk features of each candidate business; based on the current business data of the target business in the current sampling period, the target business risk features are determined from the candidate business risk features.

[0012] In one embodiment, based on current business data of the target business in the current sampling period, determining the target business risk feature from each candidate business risk feature includes:

[0013] A random forest model is used to determine the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature; based on the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature, the target business risk feature is determined from each candidate business risk feature.

[0014] In one embodiment, determining a risky business team from among candidate business teams based on target business risk characteristics and current business data includes:

[0015] The target business risk characteristics are used to update the standard anomaly detection model to obtain the current anomaly detection model; wherein the standard anomaly detection model is constructed based on the sample business risk characteristics and the actual risk information corresponding to the sample business risk characteristics; the current anomaly detection model is used to determine the risky business team from each candidate business team according to the current business data.

[0016] In one embodiment, determining a risky business team from each candidate business team based on current business data includes:

[0017] The current anomaly detection model is used to determine the team risk score of each candidate business team according to the current business characteristics corresponding to the current business data; and a risky business team is determined from the candidate business teams according to the team risk score of each candidate business team.

[0018] In one embodiment, according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business, a target management and control method of the risk business team is determined, including:

[0019] Based on historical business control records, determine the association between each standard business risk feature and each candidate control method; based on the business data corresponding to the risk business team in the current business data, determine the business features corresponding to the risk business team; use the business features corresponding to the risk business team as an index to query in the association relationship to obtain the target control method of the risk business team.

[0020] In a second aspect, the present application also provides a team business quality control device, including:

[0021] A feature determination module, used to determine the target business risk features covered in the current business data based on the current business data of the target business in the current sampling period and the risk features of each candidate business;

[0022] The team determination module is used to determine the risk business team from various candidate business teams based on the target business risk characteristics and current business data;

[0023] The method determination module is used to determine the target control method of the risk business team based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0024] The quality control module is used to control the business quality of the risk business team based on the target control method.

[0025] 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, the following steps are implemented:

[0026] Determine the target business risk characteristics covered by the current business data according to the current business data of the target business and the risk characteristics of each candidate business within the current sampling period;

[0027] Determine the risk business team from the candidate business teams based on the target business risk characteristics and current business data;

[0028] Determine the target management and control method of the risk business team based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0029] Based on the target management method, the business quality of the risk business team is controlled.

[0030] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0031] Determine the target business risk characteristics covered by the current business data according to the current business data of the target business and the risk characteristics of each candidate business within the current sampling period;

[0032] Determine the risk business team from the candidate business teams based on the target business risk characteristics and current business data;

[0033] Determine the target management and control method of the risk business team based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0034] Based on the target management method, the business quality of the risk business team is controlled.

[0035] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0036] Determine the target business risk characteristics covered by the current business data according to the current business data of the target business and the risk characteristics of each candidate business within the current sampling period;

[0037] Determine the risk business team from the candidate business teams based on the target business risk characteristics and current business data;

[0038] Determine the target management and control method of the risk business team based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0039] Based on the target management method, the business quality of the risk business team is controlled.

[0040] The above-mentioned team business quality control method, device, equipment, medium and program product, by determining the target business risk characteristics covered in the current business data based on the current business data of the target business in the current sampling period and the risk characteristics of each candidate business, and then determining the risk business team from each candidate business team based on the target business risk characteristics and the current business data; then, according to the business data corresponding to the risk business team in the current business data and each standard business risk characteristic, determine the target control method of the risk business team, and control the business quality of the risk business team based on the target control method. Compared with the related art, which can only perform business quality control on individual personnel, the above-mentioned method can ensure the comprehensiveness of business quality control by determining the risk business team from each candidate business team based on the target business risk characteristics covered in the current business data, and then determining the target control method of the risk business team based on the business data corresponding to the risk business team and each standard business risk characteristic. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] 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.

[0042] Figure 1 A schematic diagram of a process of a team business quality control method in an embodiment;

[0043] Figure 2 A schematic diagram of a process for determining risk characteristics of a target business in one embodiment;

[0044] Figure 3 A schematic diagram of a process for determining risk characteristics of a target business in another embodiment;

[0045] Figure 4 A schematic diagram of a process for determining a risk business team in one embodiment;

[0046] Figure 5 A schematic diagram of a process for determining a target control method in an embodiment;

[0047] Figure 6 is a schematic diagram of a model in an embodiment;

[0048] Figure 7 A flowchart of a team business quality control method in another embodiment;

[0049] Figure 8 It is a structural block diagram of a team business quality control device in an embodiment;

[0050] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0051] 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.

[0052] With the continuous development of the financial field, the business volume of financial-related businesses has also increased. Therefore, in order to ensure the service quality of the financial business team, business quality control methods have emerged. Existing business quality control methods generally directly use personnel positioning models to locate abnormal business personnel based on business data.

[0053] However, due to the characteristics of team business data such as small data volume, large data fluctuations, and unbalanced data distribution, the existing business quality control methods can only perform business quality control on individual personnel, and cannot achieve overall business instruction control of the team, reducing the comprehensiveness of business quality control.

[0054] Based on this, in an exemplary embodiment, Figure 1 As shown, a team business quality control method is provided, which is described by taking the method applied to a server as an example, and specifically includes the following steps:

[0055] S101, determining target business risk features covered in the current business data according to current business data of the target business in the current sampling period and risk features of each candidate business.

[0056] Among them, the so-called sampling period is the period for collecting business data, for example, it can be 1 month, and the current sampling period is the current period for collecting business data, for example, it can be this month; the target business is any financial business with team management needs, such as insurance business, etc.; the current business data is the business data of the target business collected during the current sampling period.

[0057] The so-called business risk characteristics are used to characterize the business data characteristics corresponding to the risk business. Furthermore, the candidate business risk characteristics are the business risk characteristics that may exist in the target business; the target business risk characteristics are the business risk characteristics covered by the current business data. For example, in the insurance business scenario, the candidate business risk characteristics may include but are not limited to abnormal insured customers, order cancellations, abnormal performance, abnormal policy returns and policy sum insured, collection, insurance, loan repayment, policy cancellation, claim settlement, loss of validity, entrustment and self-insurance, etc.

[0058] It is understandable that in order to ensure the reliability of data analysis, all candidate business risk features that may appear in the target business can be determined based on the historical business data of the target business; then, a feature determination model can be constructed based on each candidate business risk feature.

[0059] In one possible implementation, when a data collection event is detected, the current business data of the target business in the current sampling period can be obtained. The data collection event can be automatically initiated based on a fixed event initiation cycle. For example, the event initiation cycle can be set to 1 month, and the 2nd can be set as the initial initiation date, so that the data collection event can be automatically initiated on the 2nd of each month to obtain the current business data.

[0060] The current business data of the target business in the current sampling period can be input into a trained feature determination model, and the feature determination model outputs the target business risk features covered by the current business data based on the current business data and model parameters.

[0061] S102, determining a risk business team from among the candidate business teams according to the target business risk characteristics and current business data.

[0062] Among them, the so-called business team is a team of financial personnel that can handle the target business. For example, in the insurance business scenario, the business team can be an insurance sales team; the candidate business team is all business teams that can handle the target business; the risk business team is a business team among the candidate business teams that has problems with business quality.

[0063] In one possible implementation, the target business risk characteristics and current business data may be simultaneously input into a trained team determination model, and the team determination model outputs a risky business team from each candidate business team based on the target business risk characteristics, current business data and model parameters.

[0064] In another possible implementation, the business characteristics corresponding to the current business data can be compared with the target business risk characteristics for similarity. When there are business characteristics with higher similarity, the current business data corresponding to the business characteristics with higher similarity can be determined, and the business team associated with the current business data can be used as a risk business team.

[0065] S103, determining a target management and control method for the risk business team according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business.

[0066] Among them, the so-called abnormal business characteristics are data characteristics that can characterize abnormal business behavior; standard business risk characteristics are pre-defined business risk characteristics that have corresponding management and control methods. For example, standard business risk characteristics can be determined by experts in the field based on historical experience, or based on a large number of experiments, and this is not limited in this application. The target management and control method is a method used to control the business quality of risk business teams.

[0067] In one possible implementation, the business data corresponding to the risk business team in the current business data can be compared with each standard business risk feature respectively. If there is any matching standard business risk feature, the control method associated with the standard business risk feature will be used as the target control method for the risk business team.

[0068] S104, based on the target management method, control the business quality of the risk business team.

[0069] In one possible implementation, after determining the target management method for the risk business team, the target management method can be sent to the management end corresponding to the risk business team. The management end will manage and control the business quality of the risk business team based on the target management method, and regularly obtain the business data of the risk business team after the management and control to evaluate the management effect.

[0070] In the above-mentioned team business quality control method, the target business risk characteristics covered in the current business data are determined based on the current business data of the target business in the current sampling period and the risk characteristics of each candidate business, and then the risk business team is determined from each candidate business team based on the target business risk characteristics and the current business data; then, the target control method of the risk business team is determined based on the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business, and the business quality of the risk business team is controlled based on the target control method. Compared with the related art, which can only perform business quality control on individual personnel, the above-mentioned method can ensure the comprehensiveness of business quality control by determining the risk business team from each candidate business team based on the risk characteristics of the target business covered in the current business data, and then determining the target control method of the risk business team based on the business data corresponding to the risk business team and the risk characteristics of each standard business.

[0071] In order to ensure the accuracy of the target risk business characteristics, based on the above embodiments, in the embodiments of the present application, an optional method for determining the target business risk characteristics is provided, such as Figure 2 As shown, the specific steps include:

[0072] S201, extracting features from historical business data of the target business to obtain risk features of each candidate business.

[0073] The historical business data refers to the business data of the target business collected during the historical period.

[0074] It is understandable that the historical business data of the team as a whole is small in volume and has large fluctuations compared to personal data. Therefore, data enhancement technology can be used to process the historical business data to increase the diversity and quantity of the historical business data.

[0075] In order to ensure the comprehensiveness of historical business data, historical business data of the target business can be obtained from multiple dimensions. For example, in the insurance business scenario, sales staff data, sales agency data, policyholder data, policy data, and team data within the historical period can be used as historical business data. Among them, the data collection process can adopt an offline batch processing method. In addition, historical business data can be expanded by expanding the time range and data breadth of data acquisition. For example, by sliding the time window, extracting the historical business data within the window range, and then continuously sliding the window backward, thereby expanding the data volume of historical business data.

[0076] Furthermore, the acquired historical business data can be preprocessed. Specifically, since the data comes from different data sources, in order to ensure the accuracy, completeness, consistency and uniqueness of the data, the historical business data must first be cleaned. This mainly includes: (1) missing value processing, detecting empty values ​​in historical business data and filling them with default values; (2) noise processing, calculating abnormal values ​​and outliers in historical business data through statistical methods and removing them; (3) data deduplication, deleting duplicate data rows and columns in historical business data; (4) data consistency check, checking the correctness of data format and type in historical business data.

[0077] Subsequently, historical business data can be desensitized and encrypted: for sensitive information in the data, such as the insured's name, ID number and telephone number in insurance business scenarios, the relevant data can be replaced and encrypted to protect the privacy and security of the data.

[0078] In one possible implementation, the preprocessed historical business data may be input into a trained feature extraction model, and the feature extraction model may output each candidate business risk feature based on the historical business data and model parameters.

[0079] In another possible implementation, the basic attribute statistics, combination or conversion of the pre-processed historical business data can be performed through the expert experience and business understanding in the relevant field, and then more informative features can be extracted therefrom to obtain the risk features of each candidate business.

[0080] After obtaining each candidate business risk feature, since each candidate business risk feature has different units and scales, for example, the premium range is between 0 and 1 million, but the age range is between 0 and 100. Therefore, in order to ensure the reliability of candidate processing, each candidate business risk feature needs to be normalized, that is, the value of each candidate business risk feature is scaled proportionally so that each candidate business risk feature is within a similar scale range. In addition, text or categorical data needs to be converted into numerical data. For example, the gender of the salesperson needs to be discretized, mapping "male" to 0 and "female" to 1.

[0081] S202: Determine a target business risk feature from candidate business risk features based on current business data of the target business in a current sampling period.

[0082] In one possible implementation, the current business characteristics of the target business may be determined based on the current business data of the target business in the current sampling period; then, the candidate business risk characteristics with a high degree of similarity to the current business characteristics are used as the target business risk characteristics.

[0083] In the embodiment of the present application, by extracting features from historical business data of the target business, each candidate business risk feature is obtained, and then the target business risk feature is determined from the candidate business risk features, thereby ensuring the accuracy of the target business risk feature.

[0084] In order to ensure the accuracy of the target risk business characteristics, based on the above embodiment, in the embodiment of the present application, another optional method for determining the target business risk characteristics is provided, such as Figure 3 As shown, the specific steps include:

[0085] S301, using a random forest model to determine a similarity score between current business data of the target business in the current sampling period and each candidate business risk feature.

[0086] The similarity score is a numerical value that can characterize the similarity between business data and business risk characteristics.

[0087] In one possible implementation, the current business data of the target business within the current sampling period can be input into a random forest model constructed based on the candidate business risk features, and the random forest model outputs a similarity score between the current business data and each candidate business risk feature by measuring the number of times each candidate business risk feature is selected as a node split.

[0088] S302 : determining a target business risk feature from among the candidate business risk features according to a similarity score between current business data of the target business in a current sampling period and each candidate business risk feature.

[0089] In one possible implementation, after determining the similarity score between the current business data of the target business within the current sampling period and each candidate business risk feature, each similarity score can be arranged in order from large to small; then, a preset number of candidate business risk features arranged in front can be used as target business risk features.

[0090] Alternatively, the candidate business risk features with similarity scores greater than a preset score threshold can be directly used as target business risk features. The preset number and preset score threshold can be set manually or determined based on a large number of experiments, and this is not limited in the present application.

[0091] In an embodiment of the present application, a random forest model is used to determine the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature, and then determine the target business risk feature, which can ensure the accuracy of the target business risk feature.

[0092] In order to ensure the accuracy of the risk business team, based on the above embodiment, in the embodiment of the present application, an optional method for determining the risk business team is provided, such as Figure 4 As shown, the specific steps include:

[0093] S401, using the target business risk characteristics, updating the standard anomaly detection model to obtain the current anomaly detection model.

[0094] Among them, the standard anomaly detection model is constructed based on the sample business risk features and the actual risk information corresponding to the sample business risk features. Specifically, the sample business risk features can be selected from the candidate business risk features, and the actual risk information corresponding to the sample business risk features can be marked. Subsequently, the sample business risk features and the actual risk information corresponding to the sample business risk features are used to train the initial model to obtain the standard anomaly detection model. For example, in the insurance business scenario, the anomaly detection model needs to identify sales teams with low business quality. Therefore, the policy continuation rate can be used as an evaluation criterion to mark the actual risk information.

[0095] In this application, the initial model can be a tree-shaped machine learning model, including a decision tree, a distributed boosting library XGBoost, and a multi-layer perceptron MLP. Among them, the decision tree is a tree structure, in which each internal node represents a judgment on an attribute, each branch represents the output of a judgment result, and finally each leaf node represents a classification result.

[0096] XGBoost is an ensemble learning algorithm, an improvement of the gradient boosting tree. It builds a powerful ensemble model by combining multiple weak learners (usually decision trees). Multilayer Perceptron (MLP) is a classic feedforward artificial neural network model, which consists of input layer, hidden layer and output layer, and processes complex data patterns through nonlinear activation functions.

[0097] MLP can perform classification or regression tasks by learning the features and their relationships in the data, and has strong function approximation capabilities. In supervised learning, MLP adjusts weights through the back-propagation algorithm to minimize the error between the prediction and the actual result. MLP does not rely on manual feature engineering, automatically extracts features from the source data, has stronger generalization capabilities, and the model is more complex, so it requires more training data and computing resources.

[0098] In this application, a three-layer MLP can be used, that is, including an input layer, a hidden layer and an output layer. The hidden layer contains 100 neurons, uses an Adam optimizer, the activation function is a ramp function relu, and the learning rate is 0.001.

[0099] During model training, the prediction results of the initial model were weighted, and the weights of each model (decision tree, XGBoost, and MLP) were adjusted according to their contribution. In this way, models with better performance can occupy a larger proportion in the combination, thereby improving the overall performance.

[0100] In addition, the degree of model training can be evaluated from the following dimensions: 1) Precision, also known as positive prediction value, refers to the proportion of correct identification to the total number of samples identified as positive. 2) Recall, also known as recall rate, true positive rate TPR, that is, how many positive / positive samples are correctly identified. 3) The harmonic mean F1 of precision and recall rate, which takes into account both the accuracy and recall rate of the classification model. 4) ROC (Receiver Operating Characteristic Curve) curve, which evaluates the performance of the classifier by plotting the curve of true positive rate (TPR) against false positive rate (FPR) under different thresholds. 5) AUC (Area Under the Curve) value, that is, the area under the ROC curve. The larger the value, the better the classification performance. 6) KS value (Kolmogorov-Smirnov) is an indicator used to evaluate the risk differentiation ability of the model, which measures the difference between the cumulative distribution of good and bad samples. The larger the KS value, the stronger the risk differentiation ability of the model.

[0101] Furthermore, a cross-validation method can be used to improve the generalization ability of the model, that is, to improve the performance of the model on unknown data and reduce the overfitting problem of the model during the training process. In an embodiment of the present application, K-fold cross-validation can be used, where K is 10. That is, the sample business risk characteristics are divided into K subsets of equal size. In each round, one of the subsets is used as a test set, and the remaining K-1 subsets are combined as training sets. K training and testing are performed, and a different subset is selected as a test set each time. The final result is the average of the K test results.

[0102] After completing the above training steps, the standard anomaly detection model can be obtained.

[0103] In an optional implementation, after determining the target business risk features corresponding to the current business data, the target business risk features can be used to update the risk features in the standard anomaly detection model while retaining the model parameters to obtain the current anomaly detection model.

[0104] It is understandable that due to the data imbalance problem in the sample business risk characteristics, the low-risk sales team accounts for a large proportion of the data, while the samples of the high-risk sales team are relatively small, which makes the model ineffective during the training process. Therefore, the undersampling method can be used to reduce the number of majority class samples so that the number of positive and negative samples is more balanced. For example, the undersampling method is to randomly delete samples.

[0105] S402, using the current anomaly detection model, and according to the current business data, determining a risky business team from among the candidate business teams.

[0106] In order to improve the model processing efficiency, the current anomaly detection model can be distributed to different computing nodes for reasoning to improve the model's processing capabilities, reduce latency, and enhance scalability and fault tolerance. Among them, the model can adopt a horizontal splitting strategy, that is, the current business data obtained is split according to fixed rules, and each sub-node only processes a part of the current business data. After each sub-node is processed, the processing results are aggregated to the main node to obtain the model prediction results.

[0107] In an optional implementation, the current business data may be input into the current anomaly detection model, and the current anomaly detection model outputs a risky business team based on the current business data and model parameters.

[0108] In an embodiment of the present application, by adopting the target business risk characteristics, the standard anomaly detection model is updated, and the updated current anomaly detection model is used to determine the risky business team from each candidate business team, thereby ensuring the accuracy of the risky business team.

[0109] In order to ensure the accuracy of the risky business team, based on the above embodiments, in the embodiments of the present application, another optional method for determining the risky business team is provided. Specifically, the current anomaly detection model is used to determine the team risk score of each candidate business team according to the current business characteristics corresponding to the current business data; based on the team risk score of each candidate business team, the risky business team is determined from the candidate business teams.

[0110] Among them, the team risk score is used to characterize the risk level of the business team, wherein the higher the team risk score, the higher the risk of the business team.

[0111] In an optional implementation, the current business data may be input into the current anomaly detection model, and the current anomaly detection model may perform feature extraction processing on the current business data of each candidate business team to obtain the current business features corresponding to each candidate business team.

[0112] Subsequently, the team risk score of each candidate business team can be determined according to the degree of correlation between the current business characteristics of each candidate business team and the business risk characteristics in the model, and the candidate business teams whose team risk scores are greater than the score threshold are regarded as risky business teams.

[0113] In the embodiment of the present application, by using the current anomaly detection model to determine abnormal business characteristics, and determining the risky business team according to the business team identifier associated with the abnormal business characteristics, the accuracy of the risky business team can be guaranteed.

[0114] In order to ensure the accuracy of the target control method, based on the above embodiments, in the embodiments of the present application, an optional method for determining the target control method is provided, such as Figure 5 As shown, the specific steps include:

[0115] S501, determining the correlation between each standard business risk feature and each candidate control method based on historical business control records.

[0116] Among them, historical business control records are related records for business quality control in the historical period. Candidate control methods are methods for controlling business quality. Standard business risk characteristics are candidate business risk characteristics with associated control methods.

[0117] Understandably, reference Figure 6 As shown in the model diagram, due to the black box nature of the artificial intelligence model, it is difficult to explicitly judge its internal reasoning chain and decision-making process, and it can only be analyzed based on the input and output, which brings difficulties to the interpretability of the model.

[0118] Based on this, in an optional implementation, candidate business risk features and corresponding control methods can be extracted from historical business control records, and then the association relationship between each standard business risk feature and each candidate control method is generated.

[0119] In another optional implementation, the random forest model mentioned in the above steps can be used to determine the occurrence frequency of each candidate business risk feature based on the historical business data of the target business, and the candidate business risk features with higher occurrence frequencies are used as standard business risk features.

[0120] Subsequently, based on the actual experience of technical personnel in related fields, the control method corresponding to each standard business risk feature can be determined, and then the correlation relationship between each standard business risk feature and each candidate control method can be generated.

[0121] S502: Determine business characteristics corresponding to the risky business team according to business data corresponding to the risky business team in current business data.

[0122] In an optional implementation, the business data corresponding to the risky business team in the current business data can be input into a trained business feature extraction model, and the business feature extraction model outputs the business features corresponding to the risky business team based on the business data corresponding to the risky business team in the current business data and model parameters.

[0123] S503, using the business characteristics corresponding to the risk business team as an index, querying in the association relationship to obtain the target management and control method of the risk business team.

[0124] In an optional implementation, the business characteristics corresponding to the risk business team may be used as an index to perform a query in the association relationship; then, the target management and control method of the risk business team is determined based on the query result.

[0125] In an embodiment of the present application, by using the business characteristics corresponding to the risk business team as an index, a query is performed in the association relationship between each standard business risk characteristic and each candidate management and control method to obtain the target management and control method of the risk business team, thereby ensuring the accuracy of the target management and control method.

[0126] Figure 7 FIG. 1 is a flow chart of a method for controlling the quality of team business in another embodiment. Based on the above embodiment, this embodiment provides an optional example of a method for controlling the quality of team business. Figure 7 The specific implementation process is as follows:

[0127] S701, extracting features from historical business data of the target business to obtain risk features of each candidate business.

[0128] S702, using a random forest model to determine a similarity score between current business data of the target business in the current sampling period and each candidate business risk feature.

[0129] S703 , determining a target business risk feature from among the candidate business risk features according to a similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature.

[0130] S704, using the target business risk characteristics, updating the standard anomaly detection model to obtain the current anomaly detection model.

[0131] Among them, the standard anomaly detection model is constructed based on the sample business risk characteristics and the actual risk information corresponding to the sample business risk characteristics.

[0132] S705 , using the current anomaly detection model and according to the current business data, determining a risky business team from among the candidate business teams.

[0133] Optionally, a current anomaly detection model is used to determine a team risk score for each candidate business team based on current business features corresponding to current business data; and a risky business team is determined from among the candidate business teams based on the team risk scores of each candidate business team.

[0134] S706, determining the correlation between each standard business risk feature and each candidate control method based on historical business control records.

[0135] S707, determining business characteristics corresponding to the risky business team according to the business data corresponding to the risky business team in the current business data.

[0136] S708, using the business characteristics corresponding to the risk business team as an index, querying in the association relationship to obtain the target management and control method of the risk business team.

[0137] S709, manage the business quality of the risk business team based on the target management method.

[0138] The specific process of the above S701-S709 can be found in the description of the above method embodiment. The implementation principle and technical effect are similar and will not be repeated here.

[0139] 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.

[0140] Based on the same inventive concept, the embodiment of the present application also provides a team business quality control device for implementing the team business quality control 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 one or more team business quality control device embodiments provided below can refer to the limitations of the team business quality control method above, and will not be repeated here.

[0141] In an exemplary embodiment, Figure 8As shown, a team business quality control device 1 is provided, comprising: a feature determination module 10, a team determination module 20, a method determination module 30 and a quality control module 40, wherein:

[0142] The feature determination module 10 is used to determine the target business risk feature covered in the current business data according to the current business data of the target business in the current sampling period and the risk features of each candidate business;

[0143] A team determination module 20 is used to determine a risk business team from each candidate business team according to the target business risk characteristics and current business data;

[0144] A method determination module 30 is used to determine the target control method of the risk business team according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business;

[0145] The quality control module 40 is used to control the business quality of the risk business team based on the target control method.

[0146] In an exemplary embodiment, the feature determination module 10 includes:

[0147] A feature extraction unit is used to extract features from the historical business data of the target business to obtain risk features of each candidate business;

[0148] The feature determination unit is used to determine the target business risk feature from each candidate business risk feature based on the current business data of the target business in the current sampling period.

[0149] In an exemplary embodiment, the feature determination unit is specifically configured to:

[0150] A random forest model is used to determine the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature; based on the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature, the target business risk feature is determined from each candidate business risk feature.

[0151] In an exemplary embodiment, the team determination module 20 includes:

[0152] A model updating unit, used to update the standard anomaly detection model using the target business risk characteristics to obtain a current anomaly detection model; wherein the standard anomaly detection model is constructed based on the sample business risk characteristics and the actual risk information corresponding to the sample business risk characteristics;

[0153] The team determination unit is used to determine a risky business team from among various candidate business teams using the current anomaly detection model according to current business data.

[0154] In an exemplary embodiment, the team determination unit is specifically configured to:

[0155] The current anomaly detection model is used to determine the team risk score of each candidate business team according to the current business characteristics corresponding to the current business data; and the risky business team is determined from the candidate business teams according to the team risk score of each candidate business team.

[0156] In an exemplary embodiment, the mode determination module 30 is specifically used to:

[0157] Based on historical business control records, determine the association between each standard business risk feature and each candidate control method; based on the business data corresponding to the risk business team in the current business data, determine the business features corresponding to the risk business team; use the business features corresponding to the risk business team as an index to query in the association relationship to obtain the target control method of the risk business team.

[0158] Each module in the above-mentioned team business quality control device can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules 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 operations corresponding to each of the above modules.

[0159] 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. Fig. 9 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. Among them, 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. Among them, 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 business data. 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 method for team business quality control is implemented.

[0160] Those skilled in the art will understand that Fig. 9The 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.

[0161] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0162] In one 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 in the above-mentioned method embodiments are implemented.

[0163] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0164] It should be noted that the data involved in this application (including but not limited to business data, etc.) are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0165] 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.

[0166] 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.

[0167] 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 team business quality control method, characterized in that: The method comprises: Determine the target business risk characteristics covered by the current business data according to the current business data of the target business and the risk characteristics of each candidate business in the current sampling period; Determine a risk business team from among the candidate business teams according to the target business risk characteristics and the current business data; Determine the target management and control method of the risk business team according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business; Based on the target management and control method, the business quality of the risk business team is managed and controlled.

2. The method according to claim 1, characterized in that The step of determining the target business risk characteristics covered by the current business data according to the current business data of the target business in the current sampling period and the risk characteristics of each candidate business includes: Extracting features from historical business data of the target business to obtain risk features of each candidate business; Based on the current business data of the target business in the current sampling period, the target business risk feature is determined from the candidate business risk features.

3. The method according to claim 2, characterized in that The step of determining the target business risk feature from each candidate business risk feature based on the current business data of the target business in the current sampling period includes: A random forest model is used to determine the similarity score between the current business data of the target business and the risk characteristics of each candidate business during the current sampling period; According to the similarity score between the current business data of the target business in the current sampling period and each candidate business risk feature, the target business risk feature is determined from each candidate business risk feature.

4. The method according to claim 1, characterized in that: The step of determining a risky business team from among the candidate business teams according to the target business risk characteristics and the current business data includes: Adopting the target business risk characteristics, updating the standard anomaly detection model to obtain a current anomaly detection model; wherein the standard anomaly detection model is constructed based on the sample business risk characteristics and the actual risk information corresponding to the sample business risk characteristics; The current anomaly detection model is used to determine a risky business team from among the candidate business teams according to the current business data.

5. The method according to claim 4, characterized in that The adopting the current anomaly detection model and determining a risky business team from each candidate business team according to the current business data includes: Using the current anomaly detection model, and according to the current business characteristics corresponding to the current business data, determining the team risk score of each candidate business team; According to the team risk score of each candidate business team, a risky business team is determined from among the candidate business teams.

6. The method according to claim 1, characterized in that Determining the target management and control method of the risk business team according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business includes: Based on historical business control records, determine the correlation between each standard business risk feature and each candidate control method; Determining business characteristics corresponding to the risk business team according to business data corresponding to the risk business team in the current business data; The business characteristics corresponding to the risk business team are used as indexes to query the association relationship to obtain the target management and control method of the risk business team.

7. A team business quality control device, characterized in that: The device comprises: A feature determination module, used to determine the target business risk feature covered by the current business data according to the current business data of the target business in the current sampling period and each candidate business risk feature; A team determination module, used to determine a risky business team from among the candidate business teams according to the target business risk characteristics and the current business data; A method determination module, used to determine the target management and control method of the risk business team according to the business data corresponding to the risk business team in the current business data and the risk characteristics of each standard business; The quality control module is used to control the business quality of the risk business team based on the target control method.

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.

Images