Work order reservation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a work order reservation method and device, electronic equipment and a storage medium, and belongs to the technical field of software. The method comprises the steps of obtaining a user portrait data set and to-be-tested data, wherein the user portrait data set comprises user portrait features; determining model parameters of the prediction model based on the user portrait data set through a cross validation method; calculating a feature weight corresponding to each user portrait feature; and adopting a prediction model to predict the to-be-detected data based on the model parameters, the feature weights and the user portrait data set to obtain a target reservation mode. According to the embodiment of the invention, fixed-point classification can be carried out on single users, personalized work order reservation services can be provided for the users, and the user experience is improved.

Description

Technical Field

[0001] This application relates to the field of software technology, and in particular, to a work order reservation method and device, an electronic device, and a storage medium. Background Art

[0002] In the related art, there is a work order reservation service method. However, the existing method for determining the work order reservation method does not take into account the personalized needs of different people. The comprehensive existing problems are as follows:

[0003] 1) Differences in user preferences lead to different acceptance degrees of work order reservation methods, affecting the service efficiency of staff;

[0004] 2) The order of reservation methods is fixed, lacking flexibility and making it difficult to meet the diverse user needs;

[0005] 3) The allocation of work order reservation service resources is uneven, some efficient channels are not fully utilized and the labor cost is increased.

[0006] In summary, the technical problems existing in the related art need to be improved. Summary of the Invention

[0007] The main purpose of the embodiments of this application is to propose a work order reservation method and device, an electronic device, and a storage medium, which can classify a single user pointedly and is beneficial to providing personalized work order reservation services for users.

[0008] To achieve the above object, on the one hand, an embodiment of this application proposes a work order reservation method, and the method includes the following steps:

[0009] Obtain a user portrait data set and data to be measured, where the user portrait data set includes user portrait features, and the user portrait features at least include work order reservation methods;

[0010] Determine the model parameters of a prediction model based on the user portrait data set through a cross-validation method, where the prediction model is used to predict work order reservation methods;

[0011] Calculate the feature weight corresponding to each user portrait feature;

[0012] Use the prediction model to predict the data to be measured based on the model parameters, the feature weights, and the user portrait data set to obtain a target reservation method.

[0013] In some embodiments, the method further includes:

[0014] Perform normalization processing on the user portrait data set to obtain a normalized user portrait data set;

[0015] Delete redundant data in the normalized user portrait dataset to obtain a deleted user portrait dataset;

[0016] Perform noise filtering on the deleted user portrait dataset to obtain a filtered user portrait dataset.

[0017] In some embodiments, determining the model parameters of the prediction model based on the user portrait dataset by the cross-validation method includes:

[0018] Calculate the maximum mutual nearest neighbor value of the prediction model based on the user portrait dataset by the K-fold cross-validation method.

[0019] In some embodiments, calculating the feature weight corresponding to each user portrait feature includes:

[0020] Calculate the information entropy of the user portrait dataset;

[0021] Calculate the conditional entropy of the user portrait feature with respect to the user portrait dataset based on the information entropy;

[0022] Calculate the information gain based on the information entropy and the conditional entropy;

[0023] Construct the feature weight corresponding to the user portrait feature according to the information gain of the user portrait feature.

[0024] In some embodiments, the prediction model includes a dynamic mutual nearest neighbor detection model. Using the prediction model to predict the data to be measured based on the model parameters, the feature weights, and the user portrait dataset to obtain a target reservation method includes:

[0025] Use the dynamic mutual nearest neighbor detection model to calculate the mutual nearest neighbor relationship between the data to be measured and the user portrait dataset based on the model parameters and the feature weights;

[0026] Obtain the target reservation method according to the mutual nearest neighbor relationship.

[0027] In some embodiments, using the prediction model to predict the data to be measured based on the model parameters, the feature weights, and the user portrait dataset to obtain a target reservation method includes:

[0028] Calculate the first weighted Euclidean distance between the data to be measured and the user portrait feature based on the feature weights by the prediction model;

[0029] Determine the minimum number of nearest neighbors of the data to be measured within the model parameter range based on the first weighted Euclidean distance by the prediction model;

[0030] Determine the top minimum nearest neighbor number of the user portrait features as candidate features in the order from largest to smallest according to the first weighted Euclidean distance;

[0031] Calculate the local mean vector of the data to be measured and the candidate features through the prediction model;

[0032] Calculate the second weighted Euclidean distance between the data to be measured and the local mean vector through the prediction model;

[0033] Predict the data to be measured based on the second weighted Euclidean distance according to the nearest neighbor method through the prediction model to obtain the target appointment method.

[0034] In some embodiments, the work order appointment method includes a dispatching method, the user portrait features include user basic features and label data, and the obtaining of the user portrait data set and the data to be measured includes:

[0035] Obtain the dispatching method and the user basic features, where the user basic features include a user code;

[0036] Obtain the label data according to the user code;

[0037] Obtain the data to be measured.

[0038] To achieve the above object, another aspect of the embodiments of the present application proposes a work order appointment device, and the device includes:

[0039] A data acquisition module, configured to acquire a user portrait data set and data to be measured, where the user portrait data set includes user portrait features;

[0040] A parameter determination module, configured to determine the model parameters of the prediction model based on the user portrait data set through a cross-validation method;

[0041] A weight calculation module, configured to calculate the feature weight corresponding to each user portrait feature;

[0042] A prediction module, configured to predict the data to be measured based on the model parameters, the feature weights, and the user portrait data set by using the prediction model to obtain the target appointment method.

[0043] To achieve the above object, another aspect of the embodiments of the present application proposes an electronic device, where the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the foregoing method when executing the computer program.

[0044] To achieve the above object, another aspect of the embodiments of the present application provides a computer-readable storage medium storing a computer program, which when executed by a processor implements the method described above.

[0045] The embodiments of the present application at least include the following beneficial effects: The present application provides a work order reservation method and device, an electronic device, and a storage medium. This solution obtains a user portrait data set and data to be measured, determines the model parameters of the prediction model based on the user portrait data set through a cross-validation method, calculates the feature weights corresponding to each user portrait feature, realizes personalized configuration, and uses the prediction model to predict the data to be measured based on the model parameters, feature weights, and user portrait data set, performs fixed-point classification on a single user to obtain the target reservation method, which is beneficial to the optimal allocation of work order reservation service resources, is beneficial to providing personalized work order reservation services for users, and improves the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flowchart of the work order reservation method provided by the embodiments of the present application;

[0047] Figure 2 is a flowchart of the data preprocessing step of the work order reservation method provided by the embodiments of the present application;

[0048] Figure 3 is Figure 1 a flowchart of step S103 in

[0049] Figure 4 is Figure 1 a flowchart of step S104 in

[0050] Figure 5 is a specific implementation flowchart when the work order reservation method provided by the embodiments of the present application is applied to a work order reservation system;

[0051] Figure 6 is a schematic diagram of the access type service dispatching method and basic user characteristics provided by the embodiments of the present application;

[0052] Figure 7 is a schematic diagram of label data provided by the embodiments of the present application;

[0053] Figure 8 is a schematic diagram of the sample data set provided by the embodiments of the present application;

[0054] Figure 9 is a schematic diagram of the user age of the user portrait feature provided by the embodiments of the present application;

[0055] Figure 10 is a schematic diagram of the mutual nearest neighbor relationship provided by the embodiments of the present application;

[0056] Figure 11 It is a schematic diagram of the output form of the model prediction result provided by the embodiment of the present application;

[0057] Figure 12 It is a detailed schematic diagram of the prediction process provided by the embodiment of the present application;

[0058] Figure 13 It is a schematic diagram of the push of the reservation method provided by the embodiment of the present application;

[0059] Figure 14 It is a schematic diagram of the order assignment process provided by the embodiment of the present application;

[0060] Figure 15 It is a schematic diagram of the structure of the work order reservation device provided by the embodiment of the present application;

[0061] Figure 16 It is a schematic diagram of the hardware structure of the electronic device provided by the embodiment of the present application. Detailed implementation manners

[0062] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to 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. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the embodiments of the present application. They are only examples of devices and methods consistent with some aspects of the embodiments of the present application detailed in the appended claims.

[0063] It can be understood that the terms "first", "second", etc. used in the present application can be used in this article to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information can also be called the second information, and similarly, the second information can also be called the first information. Depending on the context, the words "if", "when" as used herein can be interpreted as "when...", "when...", or "in response to a determination".

[0064] The terms "at least one", "multiple", "each", "any one", etc. used in the present application, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any one refers to any one of the multiple.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0066] Before elaborating on the embodiments of this application in detail, some nouns and terms involved in the embodiments of this application are first explained, and the nouns and terms involved in the embodiments of this application are applicable to the following explanations.

[0067] 1) User preference features: mainly refer to user - related personal data, including user historical interaction records, behavior patterns, preference settings, etc., that is, a refined user profile.

[0068] 2) Work order reservation: Taking the dispatching of telecom work orders as an example, it mainly refers to a certain telecom user choosing a reservation method when handling access - type services. Currently, the dispatching methods for telecom access - type services are: SMS, WeChat, AI, manual, and immediate sales and reservation dispatching.

[0069] In related technologies, there is a work order reservation service method. However, the existing method for determining the work order reservation method does not consider the personalized needs of different people. The comprehensive existing problems are as follows:

[0070] 1) Differences in user preferences lead to different acceptance levels of work order reservation methods, affecting the service efficiency of staff;

[0071] 2) The order of reservation methods is fixed, lacking flexibility and difficult to adapt to the diverse needs of users;

[0072] 3) The allocation of work order reservation service resources is uneven, some efficient channels are not fully utilized and the labor cost is increased.

[0073] In summary, the technical problems existing in related technologies need to be improved.

[0074] In view of this, an embodiment of this application provides a work order reservation method, device, equipment, and medium. This solution obtains a user profile data set and data to be measured, determines the model parameters of a prediction model based on the user profile data set through a cross - validation method, calculates the feature weights corresponding to each user profile feature, realizes personalized configuration, uses the prediction model to predict the data to be measured based on the model parameters, feature weights, and user profile data set, performs fixed - point classification on a single user, and obtains the target reservation method, which is beneficial to the optimal allocation of work order reservation service resources, is beneficial to providing personalized work order reservation services for users, and improves the user experience.

[0075] The work order reservation method provided by the embodiments of the present application relates to the field of software technology. The work order reservation method provided by the embodiments of the present application can be applied to a terminal, or to a server, or can also be software running on a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, a vehicle-mounted terminal, etc., but is not limited thereto; the server side can be configured as an independent physical server, or can be configured as a server cluster or a distributed system composed of multiple physical servers, or can also be configured as a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application implementing the work order reservation method, etc., but is not limited to the above forms.

[0076] The present application can be used in many general or specific computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet-type devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and so on. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0077] It should be noted that in each specific embodiment of the present application, when it comes to performing relevant processing based on data related to the user's identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first. Moreover, the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiments of the present application need to obtain the user's sensitive personal information, the user's separate permission or separate consent will be obtained through methods such as pop-up windows or jumping to a confirmation page. After clearly obtaining the user's separate permission or separate consent, the necessary user-related data for the normal operation of the embodiments of the present application will be obtained.

[0078] Figure 1 is an optional flowchart of the work order reservation method provided by the embodiments of the present application. Figure 1The method in may include but is not limited to steps S101 to S104.

[0079] Step S101, obtain a user profile dataset and the data to be tested.

[0080] Specifically, the user profile dataset includes user profile features, and the user profile features at least include the work order reservation method.

[0081] Exemplarily, the work order reservation method includes the dispatching method, and the user profile features include user codes, dispatching methods, user basic features, and tag data.

[0082] It can be understood that the user profile dataset is the training set, and the data to be tested is the test set.

[0083] Optionally, new features (i.e., a new user profile dataset) can be constructed based on existing features (the existing user profile dataset) to enhance the expression ability of the model. Among them, new features can be constructed by combining multiple variables, transforming variable forms, etc.

[0084] In some embodiments, perform normalization processing on the user profile dataset to obtain the normalized user profile dataset; delete redundant data in the normalized user profile dataset to obtain the deleted user profile dataset; perform noise filtering processing on the deleted user profile dataset to obtain the filtered user profile dataset.

[0085] Furthermore, the user profile dataset includes the dispatching method, and the prediction model trained by the user profile dataset is used to predict the dispatching method of the user.

[0086] In some embodiments, obtain the dispatching method and user basic features; obtain tag data according to the user code; obtain the data to be tested.

[0087] In this embodiment, obtaining the user profile dataset and the data to be tested prepares the data for subsequent work order reservation.

[0088] Step S102, determine the model parameters of the prediction model based on the user profile dataset through the cross-validation method.

[0089] Specifically, the prediction model is used to predict the target work order reservation method of the data to be tested.

[0090] Optionally, adjust the parameters of the prediction model by optimizing the loss function, and use the cross-validation method (such as k-fold cross-validation) to evaluate the performance of the prediction model on different data subsets.

[0091] It is understandable that during the model training phase, cross-validation techniques can be utilized to optimize the selected training set model (i.e., the prediction model) to determine its optimal parameters (i.e., the maximum mutual nearest neighbor value), thereby enhancing the generalization ability of the model.

[0092] Optionally, the prediction model can be a dynamic nearest neighbor model, a neural network model, or a logistic regression model, among others.

[0093] In some embodiments, the maximum mutual nearest neighbor value of the prediction model is calculated based on the user profile dataset through the K-fold cross-validation method.

[0094] Among them, in the K-fold cross-validation method, the user profile dataset is placed into the model for prediction and training, thereby determining the maximum mutual nearest neighbor value of the model parameters for subsequent mutual nearest neighbor detection, while avoiding the problems of overfitting and underfitting.

[0095] In this embodiment, determining the model parameters of the prediction model based on the user profile dataset through the cross-validation method can improve the generalization ability of the model and avoid the problems of overfitting and underfitting.

[0096] Step S103, calculate the feature weight corresponding to each user profile feature.

[0097] Optionally, calculate the information entropy of the feature. The lower the entropy (less uncertainty), the higher the weight.

[0098] In some embodiments, calculate the information entropy of the user profile dataset; calculate the conditional entropy of the user profile feature with respect to the user profile dataset based on the information entropy; calculate the information gain based on the information entropy and the conditional entropy; construct the feature weight corresponding to the user profile feature according to the information gain of the user profile feature.

[0099] Optionally, the Pearson correlation coefficient can be used to measure the linear relationship between the user profile feature and the target variable, or the importance of the feature can be calculated through decision tree algorithms, or the weight can be obtained through logistic regression, among others.

[0100] In some embodiments, if the weight optimization result deviates from accuracy due to improper noise removal of the user feature data, the feature selection method will be used in parallel to calculate the weight. The feature selection method is a set of strategies for determining feature weights based on specific algorithms, including the variance selection method, the mutual information method, the model-based selection method, etc.

[0101] In this embodiment, calculating the feature weight corresponding to each user profile feature is beneficial for evaluating the importance of different user features in prediction, providing personalized work order reservation services for users, and enhancing the user experience.

[0102] Step S104: Use the prediction model to predict the data to be measured based on the model parameters, feature weights, and user profile dataset to obtain the target reservation method.

[0103] It can be understood that the prediction logic is adjusted according to the feature weights.

[0104] Optionally, set the model parameters (i.e., the optimal parameters) for the prediction model, train the prediction model through the feature weights and the user profile dataset, and use the trained prediction model to predict the data to be measured to obtain the target reservation method.

[0105] In some embodiments, according to the distribution and characteristics of the known samples, predict the work order reservation method of the data to be measured, and the dynamic mutual nearest neighbor prediction method can be adopted to complete the entire data processing and model training process.

[0106] Optionally, use the dynamic mutual nearest neighbor detection model to calculate the mutual nearest neighbor relationship between the data to be measured and the user profile dataset based on the model parameters and feature weights; obtain the target reservation method according to the mutual nearest neighbor relationship.

[0107] In some embodiments, calculate the first weighted Euclidean distance between the data to be measured and the user profile features based on the feature weights through the prediction model; determine the minimum number of nearest neighbors of the data to be measured within the model parameter range based on the first weighted Euclidean distance through the prediction model; determine the first minimum number of user profile features as candidate features in the order from largest to smallest of the first weighted Euclidean distance; calculate the local mean vector between the data to be measured and the candidate features through the prediction model; calculate the second weighted Euclidean distance between the data to be measured and the local mean vector through the prediction model; use the prediction model to predict the data to be measured based on the second weighted Euclidean distance according to the nearest neighbor method to obtain the target reservation method.

[0108] Furthermore, use the prediction model to predict the data to be measured based on the model parameters, feature weights, and user profile dataset to obtain the target dispatching method, and perform personalized dispatching services for users through the target dispatching method.

[0109] In this embodiment, using the prediction model to predict the data to be measured based on the model parameters, feature weights, and user profile dataset to obtain the target reservation method can effectively convert the user profile information into specific classification decisions.

[0110] Steps S101 to S104 shown in the embodiments of the present application, by obtaining a user portrait dataset and data to be measured, determining the model parameters of a prediction model based on the user portrait dataset through a cross-validation method, calculating the feature weights corresponding to each user portrait feature, realizing personalized configuration, using the prediction model to predict the data to be measured based on the model parameters, feature weights and the user portrait dataset, performing fixed-point classification on a single user, and obtaining a target reservation method, which is beneficial to the optimal allocation of work order reservation service resources, is beneficial to providing personalized work order reservation services for users, and improves the user experience.

[0111] Please refer to Figure 2 , in some embodiments, the work order reservation method provided by the embodiments of the present application further includes a data preprocessing step, and the data preprocessing step may include but is not limited to steps S201 to S203:

[0112] Step S201, perform normalization processing on the user portrait dataset to obtain a normalized user portrait dataset.

[0113] In some embodiments, the original user portrait dataset is obtained.

[0114] It can be understood that since the sample feature measurement values in the user portrait dataset are all different, directly using them in the K-nearest neighbor model will result in larger data weights for higher magnitudes. To calculate the weights between different feature attributes, it is necessary to first perform data normalization processing on the input data.

[0115] Among them, the essence of the normalization process is to place the value ranges of each feature in a unified numerical interval to eliminate the influence of feature differences.

[0116] Optionally, Min-Max normalization can be used, Z-Score normalization can also be used, and Robust normalization can also be used, and it is not limited to this.

[0117] In step S201 of some embodiments, the following formula can be used for normalization processing. The calculation formula for normalization processing is:

[0118]

[0119] Among them, i = 1, 2,..., m, l = 1, 2,..., m, where m represents the total number of samples (user portrait dataset), and all original data (i.e., the original user portrait dataset) will undergo normalization processing so that its value falls within the interval [0, 1].

[0120] In this embodiment, the user portrait dataset is normalized to obtain the normalized user portrait dataset, which helps the model find the optimal solution after normalization, improves the comparability of features, enhances the robustness of the model, and reduces the negative impact of outliers.

[0121] Step S202: Delete the redundant data in the normalized user portrait dataset to obtain the user portrait dataset after deletion.

[0122] It can be understood that for the goal of reducing the space complexity, the sample points in the user portrait dataset are integrated, and the user portrait dataset is streamlined by identifying and deleting those redundant sample points whose surroundings are all of the same category.

[0123] Optionally, according to the requirements of the problem, those samples that are "uncontributive" to model training are selectively removed, and the "partial distance" method is used in this process.

[0124] In step S202 of some embodiments, deleting duplicate samples is implemented through pandas.DataFrame.drop_duplicates().

[0125] Optionally, the correlation coefficient matrix between features can be calculated to delete features with too high correlation; or principal component analysis (PCA) or autoencoder can be used for dimensionality reduction, and this is not limited thereto.

[0126] In this embodiment, the redundant data in the normalized user portrait dataset is deleted to obtain the user portrait dataset after deletion, which shortens the model training time by reducing the data dimension, prevents overfitting, and improves the interpretability of the model.

[0127] Step S203: Perform noise filtering on the user portrait dataset after deletion to obtain the filtered user portrait dataset.

[0128] It can be understood that in order to ensure data quality, noise filtering is performed, aiming to screen out a more suitable sample set from the user portrait dataset and lay a solid foundation for subsequent analysis.

[0129] Optionally, the nearest neighbor sample points between the minority class sample points are used to screen and filter the noise data; then for a random sample in, the Euclidean distance to each sample in T is calculated, and when it is inconsistent with the nearest neighbor sample, this point is determined as noise and deleted.

[0130] In step S203 of some embodiments, the noise filtering model is as follows:

[0131] Input: T (i.e., the user portrait dataset after deletion), T i, A. Where A is a set of qualified samples (i.e., filtered user portrait data).

[0132] Output: T, T i .

[0133] The specific implementation process of noise filtering by the noise filtering model is as follows:

[0134] Step1 Initialize A.

[0135] Step2 Calculate the Euclidean distance between x i and each point in T, and find the point closest to x i in terms of distance.

[0136] Step3 If x and x i are not of the same category, then regard x i as a noise point, otherwise regard it as a qualified sample point and add it to A.

[0137] Step4 Update T and T according to A i .

[0138] Optionally, filtering can be performed by the 3σ principle, or by the box plot method, or by using a clustering method, etc.

[0139] In this embodiment, performing noise filtering on the user portrait data set after deletion can obtain a filtered user portrait data set, which can improve data quality, enhance model generalization, and improve prediction stability.

[0140] In the steps S201 to S203 illustrated in the embodiments of the present application, by performing normalization processing on the user portrait data set, a normalized user portrait data set is obtained, redundant data in the normalized user portrait data set is deleted to obtain a user portrait data set after deletion, and noise filtering is performed on the user portrait data set after deletion to obtain a filtered user portrait data set, realizing the process of normalization → deleting redundancy → noise filtering. Among them, normalizing first can ensure that subsequent steps are not affected by the dimension, and deleting redundant data before noise filtering can avoid misdeleting valid features due to noise.

[0141] Please refer to Figure 3 , in some embodiments, step S103 may include but is not limited to steps S301 to S304:

[0142] Step S301, calculate the information entropy of the user portrait data set.

[0143] In step S301 of some embodiments, the formula for calculating the information entropy H(T) of the user portrait data set T is:

[0144]

[0145] Among them, |C k | is the number of samples of the k-th type of feature; |T| is the sample size; k is the number of categories.

[0146] Step S302: Calculate the conditional entropy of the user portrait features with respect to the user portrait dataset based on information entropy.

[0147] In step S302 of some embodiments, taking a certain user portrait feature, user age, as an example, calculate the conditional entropy of feature B user age (i.e., the user portrait feature) with respect to dataset T:

[0148]

[0149] Among them: |T i | represents the number of samples in the sample set (i.e., the user portrait dataset) for which feature B user age (i.e., the user portrait feature) takes the value i; |C ik | represents the subset T i in which belong to class T k of the number of samples.

[0150] Step S303: Calculate the information gain based on information entropy and conditional entropy.

[0151] In step S303 of some embodiments, the information gain g(T, B) of the user portrait feature can be expressed as:

[0152] g(T, B) = H(T) - H(T|B)

[0153] Among them: H(T) is the information entropy of the user portrait dataset T, that is, the degree of uncertainty about the work order reservation method of dataset T; H(T|B) is the conditional entropy of the user portrait dataset T, that is, the degree of uncertainty of dataset T under the condition of feature B user age.

[0154] It can be understood that through the above information gain formula, the information gain of each user portrait feature in the dataset can be calculated.

[0155] Step S304: Construct the feature weights corresponding to the user portrait features according to the information gain of the user portrait features.

[0156] It can be understood that the greater the information gain, the greater the decrease in the degree of uncertainty of feature x with respect to the prediction result y, and it also means that the higher the importance of the feature with a greater information gain.

[0157] In step S304 of some embodiments, construct the feature weights according to the information gain of the features. If the user dataset T has a total of n user portrait features, then the weight of feature B user age is:

[0158]

[0159] Among them, T is the user portrait dataset, n is the number of user portrait features, the user age B is a user portrait feature, g(T,B) is the information gain of the user portrait feature B, and g(T,i) is the information gain of the i-th user portrait feature in the user portrait dataset.

[0160] It can be understood that the weight values of each feature of other user portraits can be obtained through the above weight formula, and the weight results are output and sorted to obtain the importance of different user features for work order reservation.

[0161] It should be noted that obtaining feature weights by information gain is a branch in the decision tree method. The decision tree method is a machine learning algorithm based on a tree structure, which can be used to determine the importance of features. Within the framework of this algorithm, the importance of a feature is usually measured according to the splitting frequency caused by the feature in the tree node or the information gain it contributes. Specifically, a feature is considered to have higher importance if it causes more splits or has a greater information gain in the decision tree. One of the major advantages of the decision tree method is that it can reveal the non-linear relationships in the data. However, it is sensitive to high-dimensional data and data containing noise.

[0162] Please refer to Figure 4 , in some embodiments, step S104 may include but is not limited to steps S401 to S406:

[0163] Step S401, calculate the first weighted Euclidean distance between the data to be measured and the user portrait features based on the feature weights through the prediction model.

[0164] In step S401 of some embodiments, calculate the Euclidean distance after weighting the sample information gain between the data to be measured x and the training set :

[0165]

[0166] Among them, is the first weighted Euclidean distance, x is the data to be measured, is the user portrait feature of the user portrait dataset , and w is the feature weight.

[0167] Step S402, determine the minimum number of nearest neighbors of the data to be measured within the model parameter range based on the first weighted Euclidean distance through the prediction model.

[0168] In step S402 of some embodiments, query the data to be measured x in the training set Reach mutual nearest neighbors within the maximum mutual nearest neighbor value range (i.e., the model parameter range). If sample x j Among the k-nearest neighbors of x i , that is, x j ∈N k (x i ), and at the same time, sample x i is also among the k-nearest neighbors of x j , that is, x i ∈N k (x j ), then it is said that samples x i and x j form a k-mutual nearest neighbor relationship, and calculate the minimum number of nearest neighbors that meet the condition of forming a mutual nearest neighbor relationship, denoted as

[0169] Step S403: Determine the candidate features as the first minimum number of nearest neighbor user portrait features in descending order of the first weighted Euclidean distance.

[0170] In step S403 of some embodiments, sort the first weighted Euclidean distances according to the distance, and take the first nearest neighbors for the sorting result. The expressions for the first nearest neighbors are as follows:

[0171]

[0172] where is the first nearest neighbors, is the minimum number of nearest neighbors, and x is the user portrait feature corresponding to the first weighted Euclidean distance of the first nearest neighbors.

[0173] Step S404: Calculate the local mean vector of the data to be measured and the candidate features through the prediction model.

[0174] In step S404 of some embodiments, calculate the local mean vector y of the first nearest neighbors of the data to be measured x in the training set i :

[0175]

[0176] where y i is the local mean vector, is the minimum number of nearest neighbors, is the user portrait data set The user portrait features of the first j nearest neighbors.

[0177] Step S405: Calculate the second weighted Euclidean distance between the data to be measured and the local mean vector through the prediction model.

[0178] In step S405 of some embodiments, the weighted Euclidean distance between the data x to be measured and the local mean vector y is calculated. i :

[0179]

[0180] where d i is the second weighted Euclidean distance, w is the feature weight, x is the user portrait feature, and y i is the local mean vector.

[0181] Step S406: Use the prediction model to predict the data to be measured based on the second weighted Euclidean distance according to the nearest neighbor method, and obtain the target reservation method.

[0182] In step S406 of some embodiments, the data x to be measured is predicted according to the nearest neighbor method. If the following conditions are met:

[0183] d c = arg min{d i}, i = 1, 2,..., M

[0184] where d c is the predicted class label c of the data x to be measured, and d i is the second weighted Euclidean distance.

[0185] It can be understood that the reservation method with the smallest distance will be selected as the predicted work order reservation method w of the test sample x. i . If there are multiple cases where the minimum distances d i are equal, one will be randomly selected from these corresponding reservation methods as the class label of the data x to be measured.

[0186] Taking the work order reservation system as an example, Figure 5 is a specific implementation flowchart when the work order reservation method provided in the embodiments of the present application is applied to the work order reservation system. Figure 5 The method in

[0187] Step 1: Obtain the sample data set.

[0188] Optionally, the data set source is as follows:

[0189] ① Obtain the dispatching method of access services and the basic user characteristics.

[0190] Exemplarily, the schematic diagram of the dispatching method of access services and the basic user characteristics is as Figure 6 shown.

[0191] ② Conduct insight into relevant user feature data and obtain it with the user code as the association item.

[0192] Among them, relevant user features are obtained with the user code as the associated item, and then a sample data set is obtained.

[0193] Exemplarily, the schematic diagram of the label data is as Figure 7 shown, and the schematic diagram of the sample data set is as Figure 8 shown.

[0194] Step 2, data preprocessing.

[0195] In some embodiments, sample data normalization processing is performed. Since the sample feature measurement values in the user portrait data set are all different, directly using them in the K-nearest neighbor model will result in a larger weight for data with a higher magnitude. To calculate the weights between different feature attributes, it is necessary to first perform data normalization processing on the input data, which essentially places the value ranges of each feature in a unified numerical interval to eliminate the influence of feature differences. The present invention uses the following formula for normalization processing, and the calculation formula is:

[0196]

[0197] where i = 1, 2,..., m, l = 1, 2,..., m, where m represents the total number of samples, and all original data will be normalized so that their values fall within the interval [0, 1].

[0198] In some embodiments, in order to ensure data quality, noise filtering processing is performed, aiming to screen out a more suitable sample set from the user portrait data set to lay a solid foundation for subsequent analysis. Mainly, the nearest neighbor sample points between a few category sample points are used to screen and filter the noise data; then for a random sample in, calculate the Euclidean distance to each sample in T. When it is different from the nearest neighbor sample, this point is determined as noise and deleted.

[0199] The noise filtering model is as follows:

[0200] Input: T, T i , A. Where A is the qualified sample set.

[0201] Output: T, T i .

[0202] Step1 Initialize A.

[0203] Step2 Calculate the Euclidean distance between x i and each point in T, and find the point closest to x i in distance.

[0204] Step3 If x and x i are not of the same category, then x iBe regarded as noise points, otherwise be regarded as qualified sample points and added to A.

[0205] Step4 Update T and T according to A i 。

[0206] In some embodiments, redundant samples are eliminated by "partial distance". For the goal of reducing the space complexity, the sample points in the data set are integrated, and the data set is streamlined by identifying and deleting those redundant sample points that are all of the same category around. In addition, according to the requirements of the problem, those samples that "make no contribution" to the model training are selectively eliminated, and the "partial distance" method is adopted in this process.

[0207] Step 3, determine the maximum mutual nearest neighbor value through cross-validation.

[0208] In some embodiments, in the model training stage, the cross-validation technique is used to optimize the selected training set model to determine its best parameter, the maximum mutual nearest neighbor value, so as to ensure the generalization ability of the model on the test set. The K-fold cross-validation method is adopted in the present invention. All data are placed into the model to participate in prediction and training, and accordingly the maximum mutual nearest neighbor value, the model parameter for the next mutual nearest neighbor detection, is determined, while avoiding the problems of overfitting and underfitting.

[0209] Step 4, determine the feature weights through information gain.

[0210] In some embodiments, in order to evaluate the importance degree of different user features in prediction, the method of calculating information gain is adopted to determine the weights of each user portrait feature.

[0211] Exemplarily, the schematic diagram of the user portrait feature user age is as Figure 9 shown. Taking the single feature user age as an example, let it be feature B. When the training user data set is set as T, the information gain of feature B user age for the training set T can be expressed as:

[0212] g(T,B) = H(T) - H(T|B)

[0213] Where: H(T) is the information entropy of the data set T, that is, the degree of uncertainty about the work order reservation method of the data set T; H(T|B) is the conditional entropy of the data set T, that is, the degree of uncertainty of the data set T under the condition of feature B user age.

[0214] Specifically, the steps of determining the weights of each user portrait feature by the method of calculating information gain are as follows:

[0215] Step1 Calculate the information entropy of the training data set T:

[0216]

[0217] Where: |C k | is the number of samples of the k-th type of feature; |T| is the sample size; k is the number of categories.

[0218] Step2 Calculate the conditional entropy of feature B user age with respect to the dataset T:

[0219]

[0220] Where: |T i | represents the number of samples where the value of feature B user age is i in the sample set; |C ik | represents the subset T i in the samples belonging to class T k of the number of samples.

[0221] Step3 Calculate the information gain g(T, B):

[0222] Generally speaking, the larger the information gain, the greater the reduction in the uncertainty of the feature x with respect to the prediction result y, and it also means that the more important the feature with a larger information gain. Construct feature weights according to the information gain of features. If the user dataset T has a total of n user portrait features, then the weight of feature B user age is:

[0223]

[0224] Where, T is the user portrait dataset, n is the number of user portrait features, user age B is a user portrait feature, g(T, B) is the information gain of user portrait feature B, and g(T, i) is the information gain of the i-th user portrait feature in the user portrait dataset.

[0225] Step4 Output the weight results.

[0226] It can be understood that, similarly, obtain the weight values of each feature of other user portraits, output the weight results and sort them to obtain the importance of different user features for the reservation method.

[0227] It should be noted that obtaining feature weights by information gain is a branch in the decision tree method. The decision tree method is a machine learning algorithm based on a tree structure and can be used to determine the importance of features. Within the framework of this algorithm, the importance of a feature is usually measured based on the splitting frequency caused by the feature in the tree nodes or the information gain it contributes. Specifically, a feature is considered to have higher importance if it causes more splits or has a larger information gain in the decision tree. One of the major advantages of the decision tree method is that it can reveal non-linear relationships in the data. However, it is sensitive to high-dimensional data and data containing noise.

[0228] In the present invention, if the weight optimization result deviates from accuracy due to improper noise rejection of user feature data, the feature selection method will be used in parallel to calculate the weight. The feature selection method is a set of strategies for determining feature weights based on specific algorithms, including variance selection method, mutual information method, model-based selection method, etc.

[0229] Step 5: Obtain the dynamic mutual nearest neighbor prediction result.

[0230] In some embodiments, according to the distribution and features of known samples, the dynamic mutual nearest neighbor prediction method is adopted to predict the reservation method of samples to be measured, and the entire data processing and model training process is completed.

[0231] Optionally, after determining the maximum mutual nearest neighbor value through K-fold cross-validation, the dynamic mutual nearest neighbor detection model is as follows:

[0232] Input: Sample to be measured (i.e., data to be measured) x, training set (i.e., user portrait data set) The training set is set as data under different reservation methods.

[0233] Output: Predicted reservation method (i.e., target reservation method) of sample to be measured x.

[0234] Step 1: Calculate the Euclidean distance after weighting the information gain of sample x and samples in the training set Sample information gain weighted Euclidean distance:

[0235]

[0236] Step 2: Query the minimum number of nearest neighbors when the sample to be measured x reaches the mutual nearest neighbor (if sample x within the range of the maximum mutual nearest neighbor value. That is, if x j in the k-neighborhood of x, that is, x i ∈N j (x k ), and at the same time sample x i in the k-neighborhood of x, that is, x i in x j 's k-neighborhood, that is, x i ∈N k (x j ), then it is said that sample x i , x j constitute a k-mutual nearest neighbor relationship, such as Figure 10 x1 and x2 in constitute a mutual nearest neighbor relationship). Let it be Figure 10 shown.

[0237] Step 3: Sort the calculation results in Step 1 according to the distance from high to low, and take the top nearest neighbors

[0238] Step 4 Calculate the local mean vector y of the first several nearest neighbors of the test sample x in the training set i :

[0239]

[0240] Step 5 Calculate the weighted Euclidean distance between the test sample x and the local mean vector y i :

[0241]

[0242] Step 6 Predict the test sample x according to the nearest neighbor method. If the following conditions are met:

[0243] d c = arg min{d i}, i = 1, 2, …, M

[0244] Then select the reservation method with the minimum distance as the predicted reservation method w of the test sample x i . If there are multiple cases where the minimum distances d i are equal, randomly select one from these corresponding reservation methods as the reservation method label of the test sample x.

[0245] Exemplarily, the schematic diagram of the model prediction result output form is as Figure 11 shown.

[0246] Furthermore, push back the optimal reservation method (i.e., the target reservation method), and set the optimal reservation method to the priority push. Exemplarily, the detailed schematic diagram of the prediction process is as Figure 12 shown, the schematic diagram of the reservation method push is as Figure 13 shown, and the schematic diagram of the order assignment process is as Figure 14 shown.

[0247] The embodiment of this application introduces the dynamic KNN model into the reservation information push of telecom access services to achieve personalized configuration; avoids resource waste caused by a single push method, and realizes the optimal configuration of service resources; can significantly improve the push efficiency of reservation information and the user's one-time response rate, and enhance the user's perception of telecom services.

[0248] Please refer to Figure 15 , the embodiment of this application also provides a work order reservation device, which can implement the above work order reservation method. The device includes:

[0249] A data acquisition module 1501, configured to acquire a user portrait data set and test data, where the user portrait data set includes user portrait features;

[0250] A parameter determination module 1502, configured to determine model parameters of a prediction model based on a user profile dataset through a cross-validation method;

[0251] A weight calculation module 1503, configured to calculate feature weights corresponding to each user profile feature;

[0252] A prediction module 1504, configured to use the prediction model to predict the data to be measured based on the model parameters, feature weights, and the user profile dataset, and obtain a target reservation method.

[0253] It can be understood that the content in the above method embodiments is applicable to the device embodiments of the present application. The functions specifically implemented by the device embodiments of the present application are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0254] An embodiment of the present application further provides an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the above work order reservation method is implemented. The electronic device can be any intelligent terminal including a tablet computer, an in-vehicle computer, etc.

[0255] It can be understood that the content in the above method embodiments is applicable to the device embodiments of the present application. The functions specifically implemented by the device embodiments of the present application are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0256] Please refer to Figure 16 , Figure 16 which schematically shows the hardware structure of an electronic device in another embodiment. The electronic device includes:

[0257] A processor 1601, which can be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is configured to execute relevant programs to implement the technical solutions provided by the embodiments of the present application;

[0258] A memory 1602, which can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), etc. The memory 1602 can store an operating system and other application programs. When implementing the technical solutions provided by the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1602, and the processor 1601 is called to execute the work order reservation method of the embodiments of the present application;

[0259] An input / output interface 1603 for implementing information input and output;

[0260] A communication interface 1604 for implementing communication interaction between this device and other devices, which can achieve communication through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.);

[0261] A bus 1605 for transmitting information between various components of the device (such as a processor 1601, a memory 1602, an input / output interface 1603, and a communication interface 1604);

[0262] Among them, the processor 1601, the memory 1602, the input / output interface 1603, and the communication interface 1604 achieve communication connections with each other inside the device through the bus 1605.

[0263] The embodiment of the present application also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the above work order reservation method is implemented.

[0264] It can be understood that the content in the above method embodiment is applicable to the present storage medium embodiment. The function specifically implemented by the present storage medium embodiment is the same as that of the above method embodiment, and the beneficial effects achieved are also the same as those of the above method embodiment.

[0265] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely provided relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0266] The work order reservation method, work order reservation device, electronic device, and storage medium provided by the embodiment of the present application obtain a user portrait data set and data to be measured, determine the model parameters of a prediction model based on the user portrait data set through a cross-validation method, calculate the feature weights corresponding to each user portrait feature, implement personalized configuration, use the prediction model to predict the data to be measured based on the model parameters, feature weights, and user portrait data set, perform fixed-point classification on a single user, and obtain a target reservation method, which is beneficial to the optimal configuration of work order reservation service resources, is beneficial to providing personalized work order reservation services for users, and improves the user experience.

[0267] The embodiments described in the embodiments of the present application are to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. As can be known to those skilled in the art, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0268] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than those shown in the figures, or combine certain steps, or different steps.

[0269] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0270] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, and their appropriate combinations.

[0271] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above figures are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0272] It should be understood that in this application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the relationship between associated objects and indicates that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously. Here, A and B can be singular or plural. The character " / " generally indicates an "or" relationship between the associated objects before and after. "At least one (item) of the following" or a similar expression refers to any combination of these items, including any combination of a single item or multiple items. For example, at least one (item) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0273] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the above units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms.

[0274] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0275] In addition, the functional units in each embodiment of this application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0276] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes: various media that can store programs such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0277] The preferred embodiments of the embodiments of this application have been described above with reference to the accompanying drawings, and thus do not limit the scope of rights of the embodiments of this application. Any modifications, equivalent replacements, and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of this application shall be within the scope of rights of the embodiments of this application.

Claims

1. A work order reservation method, characterized in that, The method includes the following steps: Obtain a user portrait dataset and data to be tested. The user portrait dataset includes user portrait features, and the user portrait features at least include work order reservation methods; Determine the model parameters of a prediction model based on the user portrait dataset through a cross-validation method. The prediction model is used to predict work order reservation methods; Calculate the feature weight corresponding to each of the user portrait features; Use the prediction model to predict the data to be tested based on the model parameters, the feature weights, and the user portrait dataset to obtain a target reservation method.

2. The method according to claim 1, wherein The method further includes: Perform normalization processing on the user portrait dataset to obtain a normalized user portrait dataset; Delete redundant data in the normalized user portrait dataset to obtain a user portrait dataset after deletion; Perform noise filtering processing on the user portrait dataset after the deletion to obtain a filtered user portrait dataset.

3. The method according to claim 1, wherein The step of determining the model parameters of the prediction model based on the user portrait dataset through a cross-validation method includes: Calculate the maximum mutual nearest neighbor value of the prediction model based on the user portrait dataset through a K-fold cross-validation method.

4. The method according to claim 1, wherein The step of calculating the feature weight corresponding to each of the user portrait features includes: Calculate the information entropy of the user portrait dataset; Calculate the conditional entropy of the user portrait features with respect to the user portrait dataset based on the information entropy; Calculate the information gain based on the information entropy and the conditional entropy; Construct the feature weight corresponding to the user portrait feature according to the information gain of the user portrait feature.

5. The method according to claim 1, wherein The prediction model includes a dynamic mutual nearest neighbor detection model. The step of using the prediction model to predict the data to be tested based on the model parameters, the feature weights, and the user portrait dataset to obtain a target reservation method includes: Use the dynamic mutual nearest neighbor detection model to calculate the mutual nearest neighbor relationship between the data to be tested and the user portrait dataset based on the model parameters and the feature weights; Obtain the target reservation method according to the mutual nearest neighbor relationship.

6. The method according to claim 1, wherein The step of using the prediction model to predict the data to be tested based on the model parameters, the feature weights, and the user portrait dataset to obtain a target reservation method includes: Calculate the first weighted Euclidean distance between the data to be tested and the user portrait features based on the feature weights through the prediction model; Determine the minimum number of nearest neighbors of the data to be tested within the range of the model parameters based on the first weighted Euclidean distance through the prediction model; Determine the first minimum number of the user portrait features as candidate features in descending order of the first weighted Euclidean distance; Calculate the local mean vector between the data to be tested and the candidate features through the prediction model; Calculate the second weighted Euclidean distance between the data to be tested and the local mean vector through the prediction model; Use the prediction model to predict the data to be tested based on the second weighted Euclidean distance according to the nearest neighbor method to obtain a target reservation method.

7. The method according to claim 1, characterized in that, The work order reservation method includes a dispatching method, and the user portrait features include user basic features and tag data. The obtaining of the user portrait data set and the data to be measured includes: Obtaining the dispatching method and the user basic features, where the user basic features include a user code; Obtaining the tag data according to the user code; Obtaining the data to be measured.

8. A work order reservation device, characterized in that, The device includes: A data acquisition module, configured to acquire a user portrait data set and data to be measured, where the user portrait data set includes user portrait features; A parameter determination module, configured to determine model parameters of a prediction model based on the user portrait data set through a cross-validation method; A weight calculation module, configured to calculate a feature weight corresponding to each of the user portrait features; A prediction module, configured to perform prediction on the data to be measured by using the prediction model based on the model parameters, the feature weights, and the user portrait data set to obtain a target reservation method.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images