Business network recommendation method and device based on multi-order moment, computer readable storage medium and electronic equipment

Abstract

The invention provides a business network recommendation method and device based on multi-order moment, a computer readable storage medium and electronic equipment. The method comprises the steps of obtaining historical business handling information of a user, screening multiple pieces of business network point information meeting user requirements in a preset area according to the business handling information, and determining travel time distribution and business handling time distribution of the user to each business network point; determining multi-order moments corresponding to the travel time distribution and the business handling time distribution, determining a percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, constructing a neural network model, and training the neural network model by adopting the percentile total time index and the business handling information to obtain a neural network model; obtaining a service network point recommendation model; and adopting the service network recommendation model to recommend a service network for the user to handle the service. The problem that an existing website recommendation technology is low in recommendation accuracy is solved.

Description

Technical Field

[0001] The present application relates to the technical field of business outlet recommendation. Specifically, it relates to a business outlet recommendation method, device, computer-readable storage medium, and electronic device based on multi-order moments. Background Art

[0002] Common existing outlet route recommendation methods basically consider screening out the outlet addresses that provide corresponding services according to the types of services required by customers and the distance. Such methods basically only consider the travel time of customers from the departure place to the outlet, without considering the fluctuation characteristics of the travel time between the departure place and the outlet. For example, when passing through some sections prone to traffic accidents or school sections near the time of going to school or leaving school, the risk of travel time fluctuation on this section will be significantly increased. Summary of the Invention

[0003] The main purpose of the present application is to provide a business outlet recommendation method, device, computer-readable storage medium, and electronic device based on multi-order moments, so as to at least solve the problem that the existing outlet recommendation technology for handling services does not consider the fluctuation characteristics of the travel time between the departure place and the outlet of customers, resulting in relatively low recommendation accuracy.

[0004] To achieve the above object, according to one aspect of the present application, a business outlet recommendation method based on multi-order moments is provided, including: obtaining the historical service handling information of a user, where the historical service handling information includes the type of service the user needs to handle, the location of the user, the risk tolerance index of the total service handling time, and the target service handling outlet; screening out information of multiple service outlets that meet the user's needs within a preset area according to the service handling information, and determining the travel time distribution and service handling time distribution of the user to each service outlet; respectively determining the multi-order moments corresponding to the travel time distribution and the service handling time distribution, determining the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the service handling time distribution, constructing a neural network model, and training the neural network model using the percentile total time index and service handling information to obtain a service outlet recommendation model; using the service outlet recommendation model to recommend service outlets for users who need to handle services.

[0005] Optionally, during the process of training the neural network model using the percentile total time index and service handling information, the method further includes: inputting the type of service the user needs to handle, the location of the user, the risk tolerance index, and the percentile total time index into the neural network model to obtain a model-predicted outlet; determining the loss function of the neural network model according to the model-predicted outlet and the target service handling outlet, and adjusting the model parameters of the neural network model according to the loss function.

[0006] Optionally, determining the percentile total time metric according to the multi - order moments corresponding to the travel time distribution and the business handling time distribution includes: substituting the multi - order moments of the travel time distribution into the Cornish - Fisher expansion formula to obtain the percentile travel time metric, and substituting the multi - order moments of the business handling time distribution into the Cornish - Fisher expansion formula to obtain the percentile handling time metric; determining the percentile total time metric according to the percentile travel time metric and the percentile handling time metric.

[0007] Optionally, after using the business outlet recommendation model to recommend business outlets for users with pending business, the method further includes: obtaining the actual handling result of the user with pending business, where the actual handling result includes the travel time and business handling time of the user with pending business to the recommended business outlet; performing iterative optimization processing on the business outlet recommendation model using the actual handling result.

[0008] Optionally, using the outlet recommendation model to recommend business outlets for users with pending business includes: obtaining the user information of the user with pending business, where the user information includes: the type of pending business, user location information, and the risk tolerance metric of the total business handling time; screening multiple business outlets that meet the user with pending business within a preset area according to the user information, and determining the travel time distribution and the business handling time distribution of the user with pending business to each of the business outlets; determining the percentile total time metric according to the travel time distribution and the business handling time distribution, and inputting the user information and the percentile total time metric into the business outlet recommendation model to obtain an outlet recommendation result, where the outlet recommendation result includes multiple recommended business outlets.

[0009] Optionally, after obtaining the outlet recommendation result, the method further includes: sorting the recommended business outlets in the outlet recommendation result in ascending order according to the percentile total time metric to obtain a business outlet recommendation sequence.

[0010] Optionally, adjusting the model parameters of the neural network model according to the loss function includes: adjusting the model parameters of the neural network model using a model optimization algorithm according to the loss function, where the model optimization algorithms include gradient descent algorithm, Bayesian optimization algorithm, regularization technique, and learning rate scheduling technique.

[0011] According to another aspect of the present application, there is provided a business outlet recommendation device based on multi-order moments, including: a first acquisition unit, configured to acquire the historical business handling information of a user, where the historical business handling information includes the type of business to be handled by the user, the location of the user, the risk tolerance index of the total business handling time, and the target business handling outlet; a determination unit, configured to screen a plurality of business outlets that meet the user's needs within a preset area according to the business handling information, and determine the travel time distribution and business handling time distribution of each user to each of the business outlets; a training unit, configured to respectively determine the multi-order moments corresponding to the travel time distribution and the business handling time distribution, determine the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, and construct a neural network model, and train the neural network model by using the percentile total time index and the business handling information to obtain a business outlet recommendation model; a recommendation unit, configured to recommend a business outlet for a user to handle a business by using the business outlet recommendation model.

[0012] According to still another aspect of the present application, there is provided a computer-readable storage medium, where the computer-readable storage medium includes a stored program, and when the program runs, it controls the device where the computer-readable storage medium is located to execute any one of the above-mentioned business outlet recommendation methods based on multi-order moments.

[0013] According to yet another aspect of the present application, there is provided an electronic device, including: one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include those for executing any one of the above-mentioned business outlet recommendation methods based on multi-order moments.

[0014] Applying the technical solution of the present application, historical business handling information of users is obtained, where the historical business handling information includes the type of business to be handled by the user, the location of the user, the risk tolerance index of the total handling time, and the target business handling network point; according to the business handling information, multiple business network point information that meets the user's needs within a preset area is screened, and the travel time distribution and business handling time distribution of the user to each business network point are determined; the multi-order moments corresponding to the travel time distribution and the business handling time distribution are respectively determined, the percentile total time index is determined according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, a neural network model is constructed, and the neural network model is trained by using the percentile total time index and the business handling information to obtain a business network point recommendation model; the business network point recommendation model is used to recommend business network points to users who need to handle business. It solves the problem that the existing network point recommendation technology for handling business does not consider the fluctuation characteristics of the travel time between the departure place and the network point of the customer, resulting in a relatively low recommendation accuracy. By introducing multi-order moments and determining the percentile total time index, the distribution characteristics of the travel time and business handling time of users to different business network points can be evaluated more accurately, so as to more accurately predict the total time consumption and recommend the most suitable business network points for users. Combining the training and optimization of the neural network model can continuously improve the accuracy and personalization of the recommendation, effectively reduce the user's waiting and travel time, and improve the user experience and business handling efficiency. In addition, through the feedback of the actual handling results, the iteration and optimization of the model are realized to ensure the continuous improvement and adaptability of the recommendation results, and a win-win solution is provided for users and business network points. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0016] Figure 1 The hardware structure block diagram of a mobile terminal for implementing a business network point recommendation method based on multi-order moments provided in an embodiment of this application is shown;

[0017] Figure 2 The flow schematic diagram of a business network point recommendation method based on multi-order moments provided in an embodiment of this application is shown;

[0018] Figure 3 The flow schematic diagram of a specific business network point recommendation method based on multi-order moments provided in an embodiment of this application is shown;

[0019] Figure 4 The structure block diagram of a business network point recommendation device based on multi-order moments provided in an embodiment of this application is shown.

[0020] Among them, the above-mentioned drawings include the following reference numerals:

[0021] 102, processor; 104, memory; 106, transmission device; 108, input / output device. Detailed implementation manners

[0022] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the description and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such used data may be interchanged under appropriate circumstances for the embodiments of the present application described herein. In addition, the terms "comprising" 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 necessarily 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 process, method, product or device.

[0025] For the convenience of description, some nouns or terms related to the embodiments of the present application are described below:

[0026] Moment: A moment is a set of measures of the distribution and morphological characteristics of a variable. The nth moment is defined as the integral of the nth power of a variable and its probability density function. Among them, the common first-order raw moment represents the mathematical expectation, the second-order central moment represents the variance, the third-order central moment represents the skewness, and the fourth-order central moment represents the kurtosis.

[0027] LSTM-BP Neural Network: LSTM stands for Long Short-Term Memory, which is a type of recurrent neural network designed to address the long-term dependence problem existing in general recurrent neural networks. It is a recurrent neural network with a chain structure. BP stands for Back Propagation, which is a multi-layer feedforward neural network trained according to the error backpropagation algorithm. The LSTM-BP neural network is a fusion neural network model that combines the advantages of both the above.

[0028] As introduced in the background art, the existing business outlet recommendation technology does not consider the fluctuation characteristics of the travel time between the customer's departure location and the outlet, resulting in relatively low recommendation accuracy. To solve the problem that the existing business outlet recommendation technology does not consider the fluctuation characteristics of the travel time between the customer's departure location and the outlet and has relatively low recommendation accuracy, the embodiments of the present application provide a business outlet recommendation method, device, computer-readable storage medium, and electronic device based on multi-order moments.

[0029] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention.

[0030] The method embodiments provided in the embodiments of the present application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking the operation on a mobile terminal as an example, Figure 1 is a hardware structure block diagram of a mobile terminal of a business outlet recommendation method based on multi-order moments according to an embodiment of the present invention. As Figure 1 shown, the mobile terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Among them, the above mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above mobile terminal. For example, the mobile terminal may further include more or fewer components than those shown in Figure 1 the figure, or have a different configuration from that shown in Figure 1 the figure.

[0031] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the business outlet recommendation method based on multi - order moments in the embodiments of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implements the above - mentioned method. The memory 104 may include high - speed random access memory, and may also include non - volatile memory, such as one or more magnetic storage devices, flash memory, or other non - volatile solid - state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories can be connected to the mobile terminal through a network. Examples of the above - mentioned network include, but are not limited to, the Internet, enterprise intranet, local area network, mobile communication network, and their combinations. The transmission device 106 is used to receive or send data via a network. Specific examples of the above - mentioned network may include the wireless network provided by the communication provider of the mobile terminal. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device 106 can be a Radio Frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0032] In this embodiment, a business outlet recommendation method based on multi - order moments running on a mobile terminal, a computer terminal, or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer - executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0033] Figure 2 It is a flowchart of the business outlet recommendation method based on multi - order moments according to the embodiments of the present application. As Figure 2 shown, the method includes the following steps:

[0034] Step S201, obtain the user's historical business handling information, where the above - mentioned historical business handling information includes the type of business the user needs to handle, the user's location, the risk - tolerance index of the total business handling time, and the target business handling outlet;

[0035] Specifically, the user inputs the urgency of handling their own business as a parameter, which serves as an indicator of the risk tolerance for the total business handling time. The risk tolerance indicator for the total business handling time refers to the degree of tolerance of the customer for the uncertainty (i.e., time volatility) of the total time required for business handling. This indicator reflects the personalized needs of the customer in time management, that is, the customer's acceptance of possible time extensions or delays during the business handling process. In actual business scenarios, different customers may have very different sensitivities to time and attitudes towards risks. Some customers may value efficiency more and be willing to accept a lower waiting risk; while some customers may have higher requirements for time certainty and be willing to accept a longer but more stable time within a certain range.

[0036] The risk tolerance indicator for the total business handling time enables the model to consider the specific needs of customers, provide more accurate and personalized service recommendations, thereby enhancing the customer experience and satisfaction.

[0037] Step S202: According to the above business handling information, screen multiple business outlet information that meets the user's needs within the preset area, and determine the travel time distribution and business handling time distribution of the user to each of the above business outlets.

[0038] Specifically, the preset area can be restricted to a circular area with a diameter of 5 km centered at the user's location.

[0039] Step S203: Respectively determine the multi - order moments corresponding to the above travel time distribution and the above business handling time distribution. Determine the percentile total time indicator based on the multi - order moments corresponding to the above travel time distribution and the above business handling time distribution, and construct a neural network model. Use the above percentile total time indicator and business handling information to train the neural network model to obtain a business outlet recommendation model.

[0040] The neural network model of this method can be a BP - LSTM combined neural network time - series prediction model. Among them, the BP neural network has the function of classification, and the LSTM model has the characteristics of time - series prediction. Using this model can better recommend suitable business outlets for users.

[0041] Specifically, the multi - order moments can include the first - order moment, second - order moment, third - order moment, and fourth - order moment. Among them, the first - order moment represents the time average value, the second - order moment represents the time variance, the third - order moment represents the time skewness, and the fourth - order moment represents the time kurtosis.

[0042] Step S204: Use the above business outlet recommendation model to recommend business outlets for users who need to handle business.

[0043] This method not only improves the accuracy of recommendations but also takes into account the personalized needs of users, such as risk tolerance, making the recommendations more considerate and practical. The application scenarios can be places where queuing is required for business handling, such as bank services, to help users quickly find the most suitable outlets.

[0044] Through this embodiment, historical business handling information of the user is obtained, where the historical business handling information includes the type of business the user needs to handle, the user's location, the risk tolerance index of the total business handling time, and the target business handling outlet; multiple business outlet information that meets the user's needs within a preset area is screened according to the above business handling information, and the travel time distribution and business handling time distribution of the user to each of the above business outlets are determined; the multi-order moments corresponding to the above travel time distribution and the above business handling time distribution are respectively determined, the percentile total time index is determined according to the multi-order moments corresponding to the above travel time distribution and the above business handling time distribution, a neural network model is constructed, and the above neural network model is trained using the above percentile total time index and business handling information to obtain a business outlet recommendation model; the business outlet recommendation model is used to recommend business outlets for users who need to handle business. By introducing multi-order moments and determining the percentile total time index, the distribution characteristics of the travel time and business handling time of the user to different business outlets can be evaluated more accurately, so as to more accurately predict the total time consumption and recommend the most suitable business outlet for the user. Combining the training and optimization of the neural network model can continuously improve the accuracy and personalization of the recommendation, effectively reduce the user's waiting and travel time, and improve the user experience and business handling efficiency. In addition, through the feedback of the actual handling results, the iteration and optimization of the model are realized to ensure the continuous improvement and adaptability of the recommendation results, providing a win-win solution for users and business outlets.

[0045] In the specific implementation process, during the process of training the neural network model using the above percentile total time index and business handling information, the above method further includes: inputting the type of business the user needs to handle, the user's location, the risk tolerance index, and the percentile total time index into the above neural network model to obtain a model-predicted outlet; determining the loss function of the above neural network model according to the above model-predicted outlet and the above target business handling outlet, and adjusting the model parameters of the above neural network model according to the above loss function. By adjusting the model parameters, the prediction accuracy of the model can be further improved and the recommendation error can be reduced. In bank services, this method can help users avoid long waits and improve handling efficiency.

[0046] Specifically, determining the percentile total time indicator according to the multi - order moments corresponding to the above - mentioned travel time distribution and the above - mentioned business handling time distribution includes: substituting the multi - order moments of the above - mentioned travel time distribution into the Cornish - Fisher expansion formula to obtain the percentile travel time indicator, and substituting the multi - order moments of the above - mentioned business handling time distribution into the above - mentioned Cornish - Fisher expansion formula to obtain the percentile handling time indicator; determining the above - mentioned percentile total time indicator according to the above - mentioned percentile travel time indicator and the above - mentioned percentile handling time indicator.

[0047] This method, aiming at the non - normal characteristics of the travel time and business handling time distributions, proposes to use the multi - order moments of the time distribution to recommend network routes to customers. By using advanced analysis tools in statistics, it can more accurately evaluate the time distribution and is applicable to scenarios where accurate time prediction is required in this solution.

[0048] Further, after using the above - mentioned business network recommendation model to recommend a business network for a user with a business to be handled, the above - mentioned method further includes: obtaining the actual handling result of the user with the business to be handled, where the above - mentioned actual handling result includes the travel time and business handling time of the user with the business to be handled to the recommended business network; performing iterative optimization processing on the above - mentioned business network recommendation model using the above - mentioned actual handling result.

[0049] This method can ensure the long - term effectiveness and adaptability of the recommendation model through continuous model optimization, and is applicable to dynamic environments that require continuous adjustment and optimization, such as business network recommendations during holidays or special weather conditions.

[0050] Furthermore, using the above - mentioned network recommendation model to recommend a business network for a user with a business to be handled includes: obtaining the user information of the user with the business to be handled, where the above - mentioned user information includes: the type of business to be handled, user location information, and the risk tolerance indicator for the total time of handling the business; screening multiple above - mentioned business networks that meet the user with the business to be handled within a preset area according to the above - mentioned user information, and determining the above - mentioned travel time distribution and the above - mentioned business handling time distribution of the user with the business to be handled to each of the above - mentioned business networks; determining the above - mentioned percentile total time indicator according to the above - mentioned travel time distribution and the above - mentioned business handling time distribution, and inputting the above - mentioned user information and the above - mentioned percentile total time indicator into the above - mentioned business network recommendation model to obtain a network recommendation result, where the above - mentioned network recommendation result includes multiple recommended business networks.

[0051] This method starts from the customer's perspective, considers the customer's own requirements for timeliness, and the established model can provide the network that most comprehensively meets the customer's requirements based on the customer's risk tolerance for handling time, travel route, and business handling duration, and can provide personalized recommendation services for users, improving user satisfaction, and is applicable to scenarios where personalized services are required.

[0052] Specifically, after obtaining the recommended outlet results, the above method further includes: sorting the recommended business outlets in the above recommended outlet results from low to high according to the above percentile total time index to obtain a business outlet recommendation sequence. Through this sorting, users can intuitively see the time consumption of the recommended outlets, which is convenient for users to make selections according to their own situations and is applicable to scenarios where users need to make independent decisions, such as online appointment services.

[0053] More specifically, adjusting the model parameters of the above neural network model according to the above loss function includes: adjusting the above model parameters of the above neural network model by using a model optimization algorithm according to the above loss function, where the above model optimization algorithm includes a gradient descent algorithm, a Bayesian optimization algorithm, a regularization technique, and a learning rate scheduling technique.

[0054] By using these optimization algorithms, this method can improve the training efficiency and generalization ability of the model and is applicable to scenarios that require large-scale data processing and model training, such as the outlet recommendation system of large chain enterprises.

[0055] In order to enable those skilled in the art to more clearly understand the technical solution of this application, the implementation process of the business outlet recommendation method based on multi-order moments of this application will be described in detail below in combination with specific embodiments.

[0056] This embodiment relates to a specific business outlet recommendation method based on multi-order moments, as Figure 3 shown, and specifically includes the following steps:

[0057] Step S1: Enter the business that the customer needs to handle, the customer's current location, and the risk tolerance index of the total business handling time (i.e., parameter p in the subsequent model);

[0058] Step S2: Outlet pre-screening, traverse all outlets within 5 kilometers that meet the customer's business needs;

[0059] Step S3: According to historical data, obtain the time distribution of each route from the customer's current location to each outlet during the current time period;

[0060] Step S4: According to the travel duration time distribution, calculate the first moment μ, the second moment σ 2 , the third moment S, and the fourth moment K of each distribution;

[0061] Substitute the obtained first four moments into the Cornish-Fisher expansion: Obtain the percentile travel time index of each route from the customer's current location to each outlet, where is a random variable related to skewness and kurtosis, where μ is the mean, i.e., the first moment, σ 2 is the variance, i.e., the second moment, and p is the quantile, which can be understood as the customer's risk tolerance.

[0062]

[0063] S is the skewness, i.e., the third moment, and K is the kurtosis, i.e., the fourth moment.

[0064] Step S5: When p is determined, sort in ascending order according to the percentile travel time;

[0065] Step S6: Obtain the time distribution of handling type-a m business at the network points;

[0066] Step S7: According to the time distribution of business handling duration, calculate the first moment, second moment, third moment, fourth moment, etc. of this distribution. Similarly, substitute the obtained first four moments into the Cornish-Fisher expansion formula to obtain the percentile time index for the customer to handle the corresponding business;

[0067] Step S8: Add the percentile travel time of the customer to the network point and the percentile business handling time of handling business at the current network point to obtain the percentile total time index;

[0068] Step S9: According to the customer's handling time risk tolerance and business type, establish an LSTM-BP neural network model between the time series T of the time period when handling business, {A1, A2,..., Ax} types of business type labels, and the percentile total time PTT(p) and the historical time series;

[0069] Step S10: Adjust the model parameters in Step S9 according to the prediction effect;

[0070] Step S11: Establish a business network point recommendation model for the customer's location and type-a m business types;

[0071] Step S12: Judge whether all business types have been trained;

[0072] Step S13: Complete the route recommendation considering the customer's location, waiting risk, and business type and the prediction of the business handling duration effect.

[0073] In this embodiment, aiming at the non - normal characteristics of the travel time and business handling time distribution, a method of using multi - order moments of the time distribution to recommend network routes to customers is proposed. On the basis of the traditional network point recommendation method that only considers time, the consideration of the non - normal situation of the travel time and queuing time distribution is added, and the characteristic parameters of multi - order moments are introduced, optimizing the route recommendation method for customers to network points. A network point route recommendation model based on multi - order moments is established, where the number of moments in the multi - order moments can be freely adjusted. In this embodiment, the derivation only goes up to the fourth - order moment, but it does not mean that it is only applicable to the first four - order moments. This embodiment extends from the first - order moment (average value) on the basis of the traditional average time to multi - order moments, more carefully depicting the distribution characteristics of travel time and business handling time, solving the problem that the time fluctuation characteristics of network point route recommendation are not depicted in the existing patents. From the perspective of customers, considering the needs of customers' own requirements for timeliness, the established model can provide the network point that best meets the customer's requirements comprehensively according to the customer's risk tolerance for handling time, travel route, and business handling duration.

[0074] The embodiment of the present application also provides a business network point recommendation device based on multi - order moments. It should be noted that the business network point recommendation device based on multi - order moments in the embodiment of the present application can be used to execute the method for recommending business network points based on multi - order moments provided in the embodiment of the present application. This device is used to implement the above - mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0075] The following introduces the business network point recommendation device based on multi - order moments provided in the embodiment of the present application.

[0076] Figure 4 It is a schematic diagram of the business network point recommendation device based on multi - order moments according to the embodiment of the present application. As Figure 4 shown, the device includes:

[0077] A first acquisition unit 41, configured to acquire the historical business handling information of the user, where the historical business handling information includes the type of business to be handled by the user, the location of the user, the risk tolerance index of the total business handling time, and the target business handling network point;

[0078] A determination unit 42, configured to screen multiple business network points that meet the user's needs within a preset area according to the business handling information, and determine the travel time distribution and business handling time distribution of each user to each of the above - mentioned business network points;

[0079] A training unit 43, configured to respectively determine multi - order moments corresponding to the above - mentioned travel time distribution and the above - mentioned business handling time distribution, determine a percentile total time index according to the multi - order moments corresponding to the above - mentioned travel time distribution and the above - mentioned business handling time distribution, construct a neural network model, and use the above - mentioned percentile total time index and business handling information to train the neural network model to obtain a business outlet recommendation model;

[0080] A recommendation unit 44, configured to use the above - mentioned business outlet recommendation model to recommend a business outlet for a user with a business to be handled.

[0081] In this embodiment, a first acquisition unit is configured to acquire the historical business handling information of a user. The historical business handling information includes the type of business to be handled by the user, the location of the user, a risk - tolerance index for the total business handling time, and a target business outlet for handling the business; a determination unit is configured to screen, according to the business handling information, multiple business outlets that meet the user's needs within a preset area, and determine the travel time distribution and the business handling time distribution of each user to each of the above - mentioned business outlets; a training unit is configured to respectively determine multi - order moments corresponding to the above - mentioned travel time distribution and the above - mentioned business handling time distribution, determine a percentile total time index according to the multi - order moments corresponding to the above - mentioned travel time distribution and the above - mentioned business handling time distribution, construct a neural network model, and use the above - mentioned percentile total time index and business handling information to train the neural network model to obtain a business outlet recommendation model; a recommendation unit is configured to use the above - mentioned business outlet recommendation model to recommend a business outlet for a user with a business to be handled. By introducing multi - order moments and determining the percentile total time index, the distribution characteristics of the travel time and business handling time of users to different business outlets can be evaluated more accurately, so as to more accurately predict the total time consumption and recommend the most suitable business outlet for users. Combining the training and optimization of the neural network model can continuously improve the accuracy and personalization of the recommendation, effectively reduce the user's waiting and travel time, and improve the user experience and business handling efficiency. In addition, through the feedback of the actual handling results, the iterative optimization of the model is realized to ensure the continuous improvement and adaptability of the recommendation results, providing a win - win solution for users and business outlets.

[0082] As an optional solution, the device further includes an input unit and an adjustment unit; the input unit is configured to input the type of business to be handled by the user, the location of the user, the risk - tolerance index, and the percentile total time index into the neural network model during the process of training the neural network model with the above - mentioned percentile total time index and business handling information to obtain a model - predicted outlet; the adjustment unit is configured to determine a loss function of the neural network model according to the model - predicted outlet and the target business outlet for handling the business, and adjust the model parameters of the neural network model according to the loss function.

[0083] An alternative solution is that the training unit includes a substitution module and a first determination module. The substitution module is used to substitute the multi - order moments of the above travel time distribution into the Cornish - Fisher expansion to obtain the percentile travel time index, and substitute the multi - order moments of the above business handling time distribution into the above Cornish - Fisher expansion to obtain the percentile handling time index. The first determination module is used to determine the above percentile total time index according to the above percentile travel time index and the above percentile handling time index.

[0084] An alternative solution is that the device further includes a second acquisition unit and an iterative optimization unit. The second acquisition unit is configured to acquire the actual handling result of the user with the business to be handled after using the above business outlet recommendation model to recommend a business outlet for the user with the business to be handled. The actual handling result includes the travel time of the user with the business to be handled to the recommended business outlet and the business handling time. The iterative optimization unit is used to perform iterative optimization processing on the above business outlet recommendation model by using the above actual handling result.

[0085] An alternative solution is that the recommendation unit includes a first acquisition module, a second determination module, and a third determination module. The first acquisition module is used to acquire the user information of the user with the business to be handled. The user information includes: the type of the business to be handled, the user location information, and the risk tolerance index of the total business handling time. The second determination module is used to screen multiple above business outlets that meet the above user with the business to be handled within a preset area according to the above user information, and determine the above travel time distribution and the above business handling time distribution of the above user with the business to be handled to each of the above business outlets. The third determination module is used to determine the above percentile total time index according to the above travel time distribution and the above business handling time distribution, input the above user information and the above percentile total time index into the above business outlet recommendation model, and obtain a business outlet recommendation result. The business outlet recommendation result includes multiple recommended business outlets.

[0086] An alternative solution is that the recommendation unit further includes a sorting module, which is used to sort the above recommended business outlets in the business outlet recommendation result in ascending order according to the above percentile total time index after obtaining the business outlet recommendation result, so as to obtain a business outlet recommendation sequence.

[0087] An alternative solution is that the adjustment unit includes an adjustment module, which is used to adjust the above model parameters of the above neural network model according to the above loss function by using a model optimization algorithm. The model optimization algorithm includes a gradient descent algorithm, a Bayesian optimization algorithm, a regularization technique, and a learning rate scheduling technique.

[0088] The above business outlet recommendation device based on multi - order moments includes a processor and a memory. The first acquisition unit, determination unit, training unit, recommendation unit, etc. are all stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The above modules are all located in the same processor; or, the above - mentioned each module is located in different processors in any combination form.

[0089] The processor contains a kernel, and the kernel retrieves the corresponding program units from the memory. One or more kernels can be set. By adjusting the kernel parameters, the problem that the existing business outlet recommendation technology for handling business does not consider the fluctuation characteristics of the travel time between the departure place of the customer and the business outlet, resulting in a relatively low recommendation accuracy, can be solved.

[0090] The memory may include non - permanent memory in a computer - readable medium, forms such as random access memory (RAM) and / or non - volatile memory, such as read - only memory (ROM) or flash memory (flash RAM). The memory includes at least one storage chip.

[0091] An embodiment of the present invention provides a computer - readable storage medium. The computer - readable storage medium includes a stored program. Wherein, when the above program runs, it controls the device where the computer - readable storage medium is located to execute the above - mentioned business outlet recommendation method based on multi - order moments.

[0092] Specifically, the business outlet recommendation method based on multi - order moments includes:

[0093] Step S201: Obtain the historical business handling information of the user. Wherein, the above - mentioned historical business handling information includes the type of business to be handled by the user, the location of the user, the risk - tolerance index of the total business handling time, and the target business handling outlet.

[0094] Step S202: Screen the information of multiple business outlets that meet the user's needs within a preset area according to the above - mentioned business handling information, and determine the travel - time distribution and business - handling - time distribution of the user to each of the above - mentioned business outlets.

[0095] Step S203: Respectively determine the multi - order moments corresponding to the above - mentioned travel - time distribution and the above - mentioned business - handling - time distribution, determine the percentile total - time index according to the multi - order moments corresponding to the above - mentioned travel - time distribution and the above - mentioned business - handling - time distribution, and construct a neural network model. Use the above - mentioned percentile total - time index and business handling information to train the neural network model to obtain a business outlet recommendation model.

[0096] Step S204: Use the above - mentioned business outlet recommendation model to recommend business outlets for users who need to handle business.

[0097] An embodiment of the present invention provides a processor, which is used to run a program. When the program runs, it executes the business outlet recommendation method based on multi-order moments.

[0098] Specifically, the business outlet recommendation method based on multi-order moments includes:

[0099] Step S201: Obtain the user's historical business handling information, where the historical business handling information includes the type of business to be handled by the user, the user's location, the risk tolerance index of the total business handling time, and the target business outlet for handling the business;

[0100] Step S202: Screen the information of multiple business outlets that meet the user's needs within a preset area according to the business handling information, and determine the travel time distribution and business handling time distribution of the user to each of the business outlets;

[0101] Step S203: Respectively determine the multi-order moments corresponding to the travel time distribution and the business handling time distribution, determine the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, construct a neural network model, and use the percentile total time index and the business handling information to train the neural network model to obtain a business outlet recommendation model;

[0102] Step S204: Use the business outlet recommendation model to recommend business outlets for users who need to handle business.

[0103] An embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements at least the following steps:

[0104] Step S201: Obtain the user's historical business handling information, where the historical business handling information includes the type of business to be handled by the user, the user's location, the risk tolerance index of the total business handling time, and the target business outlet for handling the business;

[0105] Step S202: Screen the information of multiple business outlets that meet the user's needs within a preset area according to the business handling information, and determine the travel time distribution and business handling time distribution of the user to each of the business outlets;

[0106] Step S203: Respectively determine the multi-order moments corresponding to the travel time distribution and the business handling time distribution, determine the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, construct a neural network model, and use the percentile total time index and the business handling information to train the neural network model to obtain a business outlet recommendation model;

[0107] Step S204: Recommend service outlets for users who need to handle services by using the above service outlet recommendation model.

[0108] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0109] This application also provides a computer program product. When executed on a data processing device, it is adapted to execute a program initialized with at least the following method steps:

[0110] Step S201: Obtain the historical service handling information of the user. Among them, the above historical service handling information includes the service type that the user needs to handle, the location of the user, the risk tolerance index of the total service handling time, and the target service outlet for handling services;

[0111] Step S202: Screen the information of multiple service outlets that meet the user's needs within the preset area according to the above service handling information, and determine the travel time distribution and service handling time distribution of the user to each of the above service outlets;

[0112] Step S203: Determine the multi-order moments corresponding to the above travel time distribution and the above service handling time distribution respectively. Determine the percentile total time index according to the multi-order moments corresponding to the above travel time distribution and the above service handling time distribution, and construct a neural network model. Use the above percentile total time index and service handling information to train the neural network model to obtain a service outlet recommendation model;

[0113] Step S204: Recommend service outlets for users who need to handle services by using the above service outlet recommendation model.

[0114] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. They can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules respectively, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0115] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0116] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0117] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implement the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0119] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0120] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

[0121] A computer-readable medium includes permanent and non-permanent, removable and non-removable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information accessible by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0122] It should also be noted that the term "comprises," "comprising," or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but also other elements not expressly listed, or elements that are inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0123] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0124] 1) A business outlet recommendation method based on multi - order moments in this application includes: obtaining the user's historical business handling information, where the historical business handling information includes the type of business the user needs to handle, the user's location, the risk - tolerance index of the total business handling time, and the target business outlet for handling the business; screening information of multiple business outlets that meet the user's needs within a preset area according to the business handling information, and determining the travel - time distribution and business - handling - time distribution from the user to each business outlet; respectively determining the multi - order moments corresponding to the travel - time distribution and the business - handling - time distribution, determining the percentile total - time index according to the multi - order moments corresponding to the travel - time distribution and the business - handling - time distribution, constructing a neural network model, training the neural network model using the percentile total - time index and the business handling information, and obtaining a business outlet recommendation model; using the business outlet recommendation model to recommend business outlets for users who need to handle business. By introducing multi - order moments and determining the percentile total - time index, it is possible to more accurately evaluate the distribution characteristics of the travel time and business - handling time from the user to different business outlets, thereby more accurately predicting the total time consumption and recommending the most suitable business outlets for users. Combining the training and optimization of the neural network model can continuously improve the accuracy and personalization of the recommendation, effectively reduce the user's waiting and travel time, and improve the user experience and business - handling efficiency. In addition, through the feedback of the actual handling results, the iterative optimization of the model is realized to ensure the continuous improvement and adaptability of the recommendation results, providing a win - win solution for users and business outlets.

[0125] 2) A business outlet recommendation device based on multi - order moments of the present application includes: a first acquisition unit for acquiring the user's historical business handling information, where the historical business handling information includes the type of business to be handled by the user, the user's location, the risk - tolerance index of the total handling time, and the target business outlet for handling the business; a determination unit for screening multiple business outlets that meet the user's needs within a preset area according to the business handling information, and determining the travel - time distribution and business - handling - time distribution of each user to each business outlet; a training unit for respectively determining the multi - order moments corresponding to the travel - time distribution and the business - handling - time distribution, determining the percentile total - time index according to the multi - order moments corresponding to the travel - time distribution and the business - handling - time distribution, constructing a neural network model, and training the neural network model using the percentile total - time index and the business handling information to obtain a business outlet recommendation model; a recommendation unit for recommending a business outlet for a user with a business to be handled using the business outlet recommendation model. By introducing multi - order moments and determining the percentile total - time index, it is possible to more accurately evaluate the distribution characteristics of the travel time and business - handling time of users to different business outlets, so as to more accurately predict the total time consumption and recommend the most suitable business outlet for users. Combining the training and optimization of the neural network model can continuously improve the accuracy and personalization of the recommendation, effectively reduce the user's waiting and travel time, and improve the user experience and business - handling efficiency. In addition, through the feedback of the actual handling results, the iteration and optimization of the model are realized to ensure the continuous improvement and adaptability of the recommendation results, providing a win - win solution for users and business outlets.

[0126] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A business network recommendation method based on multi-order moments, characterized in that: include: Obtaining the user's historical business handling information, wherein the historical business handling information includes the type of business that the user needs to handle, the user's location, the risk tolerance index of the total time of handling the business, and the target business handling outlet; Filtering information of multiple business outlets in a preset area that meet the needs of the user according to the business handling information, and determining the travel time distribution and business handling time distribution of the user to each of the business outlets; Determine the multi-order moments corresponding to the travel time distribution and the business handling time distribution respectively, determine the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, and construct a neural network model, and use the percentile total time index and business handling information to train the neural network model to obtain a business outlet recommendation model; The business outlet recommendation model is used to recommend business outlets to users who are waiting to handle business.

2. The method according to claim 1, characterized in that In the process of training the neural network model using the percentile total time index and the business handling information, the method further includes: Inputting the type of business to be handled by the user, the location of the user, the risk tolerance index and the percentile total time index into the neural network model to obtain the model prediction outlets; The loss function of the neural network model is determined according to the model prediction outlets and the target business handling outlets, and the model parameters of the neural network model are adjusted according to the loss function.

3. The method according to claim 1, characterized in that Determining a percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution includes: Substituting the multi-order moments of the travel time distribution into the Cornish-Fisher expansion to obtain the percentile travel time index, and substituting the multi-order moments of the business handling time distribution into the Cornish-Fisher expansion to obtain the percentile handling time index; The percentile total time index is determined based on the percentile travel time index and the percentile handling time index.

4. The method according to claim 1, characterized in that: After the service outlet recommendation model is used to recommend a service outlet for the user to handle the service, the method further includes: Obtaining the actual handling result of the user of the pending service, wherein the actual handling result includes the travel time of the user of the pending service to the recommended service outlet and the service handling time; The actual handling results are used to iteratively optimize the business outlet recommendation model.

5. The method according to claim 1, characterized in that The network point recommendation model is used to recommend business network points to users who want to handle business, including: Acquire user information of the user to be processed, wherein the user information includes: the type of the pending service, user location information and the risk tolerance index of the total time of the processing service; According to the user information, multiple business outlets in a preset area that meet the requirements of the user to be processed are screened, and the travel time distribution and business processing time distribution of the user to be processed to each business outlet are determined; The percentile total time index is determined according to the travel time distribution and the business handling time distribution, and the user information and the percentile total time index are input into the business outlet recommendation model to obtain an outlet recommendation result, which includes multiple recommended business outlets.

6. The method according to claim 5, characterized in that After obtaining the network point recommendation result, the method further includes: The recommended business outlets in the outlet recommendation result are sorted from low to high according to the percentile total time index to obtain a business outlet recommendation sequence.

7. The method according to claim 2, characterized in that Adjusting the model parameters of the neural network model according to the loss function includes: According to the loss function, a model optimization algorithm is used to adjust the model parameters of the neural network model, wherein the model optimization algorithm includes a gradient descent algorithm, a Bayesian optimization algorithm, a regularization technique, and a learning rate scheduling technique.

8. A device for recommending business outlets based on multi-order moments, characterized in that: include: A first acquisition unit is used to acquire the user's historical business handling information, wherein the historical business handling information includes the type of business that the user needs to handle, the user's location, the risk tolerance index of the total time of handling the business, and the target business handling outlet; A determination unit, configured to screen a plurality of business outlets in a preset area that meet user needs according to the business handling information, and determine the travel time distribution and business handling time distribution of each user to each of the business outlets; A training unit, used to respectively determine the multi-order moments corresponding to the travel time distribution and the business handling time distribution, determine the percentile total time index according to the multi-order moments corresponding to the travel time distribution and the business handling time distribution, and construct a neural network model, and train the neural network model using the percentile total time index and business handling information to obtain a business outlet recommendation model; The recommendation unit is used to recommend business outlets to users who want to handle business by adopting the business outlet recommendation model.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the service network recommendation method based on multi-order moments as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the multi-order moment-based business outlet recommendation method described in any one of claims 1 to 7.

Images