Financial product sending optimization method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of intelligent decision and financial science and technology, and discloses a financial product sending optimization method and device, computer equipment and a storage medium, and the method comprises the steps: screening the financial user behavior data of a financial product user according to the historical consumption behavior data of a non-financial product; performing model training on a preset classification model according to the financial user behavior data to obtain a user classification model; performing user classification on the target financial product users by using the user classification model to obtain user categories; generating a financial product sending scheme according to the user category, and obtaining sending feedback data corresponding to the financial product sending scheme; and optimizing the financial product sending scheme according to the sending feedback data to obtain a target sending scheme. And a target sending scheme which better meets the demand of the target user is generated, so that the sending accuracy of the financial product is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent decision-making and financial technology, and in particular to a method, device, computer equipment and storage medium for optimizing the delivery of financial products. Background Art

[0002] As a rapidly growing sub-type of financial product in the insurance field, pet insurance, logistics line insurance and other types of insurance have become an important guarantee for promoting general consumption. However, the target customer groups of these emerging insurance products are relatively clear and have strong personalized needs. Traditional extensive marketing strategies are difficult to meet the needs of market segmentation.

[0003] In addition, under the wave of digitalization, product recommendation and matching based on big data analysis technology has gradually become a key factor driving innovation and development in all walks of life. The e-commerce industry has accumulated a large amount of data on consumer shopping habits, preferences and consumption capacity, while the insurance industry has information on users' insurance needs, purchase history, etc. The data between the two are highly complementary, but due to industry barriers and privacy concerns, data is difficult to circulate and share.

[0004] Therefore, how to effectively conduct correlation analysis and utilization of user e-commerce consumption behavior data and insurance product consumption behavior data legally obtained with the user's knowledge and consent, while protecting user privacy, and thereby improve the accuracy of financial product recommendation matching, has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The present invention provides an artificial intelligence financial product delivery optimization method, device, computer equipment and storage medium to solve the technical problem that the data resource is single and the user cannot be accurately positioned, which leads to poor accuracy in the delivery of financial products.

[0006] In a first aspect, a method for optimizing the delivery of financial products is provided, comprising:

[0007] Acquire historical consumption behavior data of non-financial products, and filter financial user behavior data of financial product users based on the historical consumption behavior data;

[0008] Using the financial user behavior data to perform model training on a preset classification model to obtain a user classification model;

[0009] Using the user classification model to classify target financial product users to obtain user categories;

[0010] generating a financial product delivery plan according to the user category, delivering the financial product to the target financial product user according to the financial product delivery plan, and obtaining delivery feedback data corresponding to the financial product delivery plan;

[0011] The financial product delivery plan is optimized according to the delivery feedback data to obtain a target delivery plan.

[0012] In a second aspect, a financial product delivery optimization device is provided, comprising:

[0013] A data screening module, used to obtain historical consumption behavior data of non-financial products, and screen financial user behavior data of financial product users based on the historical consumption behavior data;

[0014] A model training module, used to perform model training on a preset classification model using the financial user behavior data to obtain a user classification model;

[0015] A user classification module, used to classify target financial product users using the user classification model to obtain user categories;

[0016] A product delivery plan generating module, configured to generate a financial product delivery plan according to the user category, deliver the financial product to the target financial product user according to the financial product delivery plan, and obtain delivery feedback data corresponding to the financial product delivery plan;

[0017] The product delivery plan optimization module is used to optimize the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan.

[0018] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned financial product delivery optimization method when executing the computer program.

[0019] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned financial product delivery optimization method are implemented.

[0020] In the scheme implemented by the above-mentioned financial product delivery optimization method, device, computer equipment and storage medium, historical consumption behavior data of non-financial products can be received through the client, and financial user behavior data of financial product users can be screened out based on the historical consumption behavior data; a preset classification model can be trained based on the financial user behavior data to obtain a user classification model; the user classification model can be used to classify target financial product users to obtain user categories; a financial product delivery plan is generated based on the user category, and financial products are delivered to the target financial product users based on the financial product delivery plan, and delivery feedback data corresponding to the financial product delivery plan is obtained; the financial product delivery plan is optimized based on the delivery feedback data to obtain a target delivery plan. The target sending plan is fed back to the client. In the present invention, for complex insurance entities under the insurance business, the e-commerce consumption behavior data of the e-commerce users of the financial product users and the insurance product consumption behavior data can be effectively associated by screening the financial user behavior data of the financial product users; then the preset classification model can be trained according to the financial user behavior data, and external data resources can be effectively used to improve the richness of user data and obtain a more accurate user classification model; the user classification model is then used to classify the target financial product users, generate a financial product sending plan, and obtain corresponding sending feedback data, and the financial product sending plan can be optimized according to the sending feedback data to form a closed-loop iteration, generate a target sending plan that better meets the needs of the target users, and effectively improve the accuracy of financial product recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0022] Figure 1 This is a schematic diagram of an application environment of a financial product delivery optimization method in an embodiment of the present invention;

[0023] Figure 2 is a flow chart of a method for optimizing the delivery of financial products in one embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of a process for screening financial user behavior data of financial product users provided by an embodiment of the present invention;

[0025] Figure 4 is a structural diagram of a financial product delivery optimization device in one embodiment of the present invention;

[0026] Figure 5 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0027] Figure 6 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The financial product delivery optimization method provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In an application environment, the client communicates with the server through a network. The server can receive historical consumption behavior data of non-financial products through the client, filter out financial user behavior data of financial product users according to the historical consumption behavior data; perform model training on a preset classification model according to the financial user behavior data to obtain a user classification model; perform user classification on target financial product users using the user classification model to obtain user categories; generate a financial product delivery plan according to the user category, deliver financial products to the target financial product users according to the financial product delivery plan, and obtain delivery feedback data corresponding to the financial product delivery plan; optimize the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan. Feedback the target delivery plan to the client. In the present invention, for complex insurance entities such as insurance business, by screening out the financial user behavior data of financial product users, the e-commerce consumption behavior data of financial product users and the insurance product consumption behavior data can be effectively associated; then the preset classification model can be trained according to the financial user behavior data, effectively utilizing external data resources, improving the richness of user data, and obtaining a more accurate user classification model; then the user classification model is used to classify the target financial product users, generate a financial product delivery plan, and obtain the corresponding delivery feedback data, the financial product delivery plan can be optimized according to the delivery feedback data, forming a closed-loop iteration, generating a target delivery plan that is more in line with the needs of the target user, and effectively improving the accuracy of financial product recommendations. Among them, the client can be but not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices. The server can be implemented with an independent server or a server cluster composed of multiple servers. The present invention is described in detail below through specific embodiments.

[0030] See also Figure 2 As shown, Figure 2 A flowchart of a method for optimizing the delivery of financial products provided by an embodiment of the present invention includes the following steps:

[0031] S1. Obtain historical consumption behavior data of non-financial products, and filter financial user behavior data of financial product users based on the historical consumption behavior data.

[0032] In the embodiments of the present invention, non-financial products are products that do not directly involve financial transactions or financial assets. In contrast to financial products, which can be stocks, bonds, currencies, derivatives, etc., and usually involve the lending, investment or risk management of funds, non-financial products can be physical goods such as food, beverages, clothing, furniture, electronic products, and service products such as education, medical care, travel, consulting, and entertainment.

[0033] In detail, historical consumption behavior data refers to the consumption data of non-financial products generated by users on e-commerce platforms within a preset time period, such as personal identifiers, behavior patterns, shopping habits, shopping preferences and consumption capacity data. However, historical consumption behavior data is private data for users, and e-commerce platforms also need to protect the privacy of consumers. Therefore, it is necessary to ensure that even if the product user data is intercepted during transmission, the original user personal information cannot be restored.

[0034] In the embodiment of the present invention, refer to Figure 2 As shown, the filtering of the financial user behavior data of the financial product user according to the historical consumption behavior data includes:

[0035] S21, performing de-identification processing on the historical consumption behavior data to obtain unidentified data;

[0036] S22, performing joint encryption calculation on the unidentified data to obtain joint user behavior data;

[0037] S23. Extracting financial user behavior data from the joint user behavior data.

[0038] Specifically, user identification is an attribute in historical consumption behavior data that can directly or indirectly identify the subject of personal information, including direct identifiers that can independently identify the subject of personal information in a specific environment, such as name, ID number, passport number, driver's license number, address, email address, telephone number, fax number, bank card number, license plate number, vehicle identification number, social insurance number, health card number, medical record number, device identifier, biometric code, Internet Protocol (IP) address number and network universal resource locator (URL), etc.; and quasi-identifiers that may not be sufficient to identify the subject of personal information alone, but when combined with other attributes, can uniquely identify the subject of personal information, such as gender, date of birth or age, event date (such as admission, surgery, discharge, visit), location (such as postal code, building name, region), ethnic origin, country of birth, language, indigenous identity, visible minority status, occupation, marital status, education level, years of schooling, criminal history, total income and religious beliefs.

[0039] Furthermore, de-identification refers to removing user identifiers from historical consumption behavior data, which can protect the data privacy and data security of historical consumption behavior data. Specifically, historical consumption behavior data can be de-identified by replacing, invalidating, or reordering user identifiers.

[0040] In an embodiment of the present invention, joint encryption computing is that the financial platform and the e-commerce platform establish a connection on a secure computing platform, and through a multi-party secure computing protocol, the historical consumption behavior data of users in the financial platform on the e-commerce platform are counted. Among them, the multi-party secure computing protocol involves complex cryptographic technology and protocol technology. The financial platform and the e-commerce platform first generate cryptographic materials, and then execute the set calculation steps according to the selected encryption computing protocol, so that the financial platform can obtain the historical consumption behavior data of the e-commerce platform under the premise of ensuring data security. The e-commerce platform can extract the data of users of the e-commerce platform in the historical consumption behavior data to obtain financial user behavior data.

[0041] In detail, homomorphic encryption can be used for joint encryption calculations, and then historical consumer behavior data can be safely outsourced to financial platforms for processing without worrying about the risk of data leakage. Among them, homomorphic encryption is an important means of protecting information security. It converts original data (unidentified data) into ciphertext that cannot be directly read through cryptographic principles to ensure the security of data during transmission or storage.

[0042] In detail, homomorphic encryption is a special encryption method that allows calculations to be performed directly on unidentified data without first decrypting the unidentified data. The core feature of this encryption method is that it can achieve ciphertext homomorphic operations, that is, performing specific calculation operations on ciphertext, and the ciphertext calculation results obtained after decryption are equivalent to the same calculations performed directly on the plaintext data, which allows the owner of the data to safely outsource the data to a third party for processing without having to worry about the risk of data leakage.

[0043] In an embodiment of the present invention, by screening the financial user behavior data of financial product users from the historical consumption behavior data, the historical consumption behavior data of non-financial products on the e-commerce platform can be combined with the financial product users while ensuring data privacy, effectively complementing the data and improving the security and comprehensiveness of the data.

[0044] S2. Using the financial user behavior data, a preset classification model is trained to obtain a user classification model.

[0045] In an embodiment of the present invention, in a financial platform such as the financial insurance industry, the insurance industry has information such as the user's insurance needs and purchase history, which is highly complementary to the shopping habits, preferences and consumption capacity data in the historical consumption behavior data. Therefore, for the insurance industry, it is possible to comprehensively consider the user's product user data on the e-commerce platform and the user data in the insurance industry, classify users, identify potential insurance purchasing groups, and design personalized product recommendations and marketing activities to improve the pertinence and effectiveness of marketing.

[0046] In an embodiment of the present invention, the financial platform can establish a secure connection with the e-commerce platform, and transmit the user encrypted data and encryption key to the financial platform. The financial platform then decrypts the encrypted data based on the key pair. If the financial platform is secure, it has the correct key to decrypt the encrypted data, and then screen out the financial user behavior data of the users of the financial product to ensure the security of the historical consumption behavior data.

[0047] Furthermore, the preset classification model may be a machine learning model, such as a logistic regression model, a support vector set model, an ensemble learning model, a neural network model, etc., and a user classification model may be obtained by optimizing model parameters in the machine learning model.

[0048] In the embodiment of the present invention, the method of using the financial user behavior data to perform model training on a preset classification model to obtain a user classification model includes:

[0049] Using the classification model to classify the financial user behavior data to obtain a predicted category;

[0050] Calculating the loss value of the classification model according to the predicted category, optimizing the parameters of the classification model when the loss value is greater than a preset loss value threshold, and returning to the step of using the classification model to classify the financial user behavior data to obtain the predicted category;

[0051] When the loss value is less than the loss value threshold, a user classification model is obtained.

[0052] Specifically, the historical consumption behavior data of financial users on e-commerce platforms can be extracted from the insurance company's user database, including corresponding insurance search records and consumption behaviors, so as to design targeted insurance product recommendations and marketing activities.

[0053] Furthermore, the classification model can be used to classify users in the financial user behavior data into users with different purchasing tendencies, which can be used to identify potential insurance product purchasers. For example, users can be classified into high-potential, medium-potential, and low-potential product purchasers, thereby obtaining different prediction categories.

[0054] In an embodiment of the present invention, the financial user behavior data includes the actual purchasing tendency of the financial user in the financial platform, that is, the actual user category of each financial user, and the actual user category and the predicted category are calculated using a preset model loss function to obtain the loss value of the classification model. Among them, the loss function is a function that measures the difference between the model prediction value and the actual value, that is, the function that calculates the difference between the actual user category and the predicted category in the present invention, which can be a mean square error (MSE) loss function, a mean absolute error (MAE) loss function, a cross-entropy loss function, and a logarithmic loss (Log Loss) function, etc. The prediction error of the classification model can be quantified through the loss function, thereby improving the prediction performance of the classification model and obtaining a more accurate user classification model.

[0055] Among them, the Mean Squared Error (MSE) loss function is to calculate the difference between each pair of actual user categories and predicted categories, that is, the residual, and square each residual to eliminate the influence of negative values, ensuring that all deviations are positive values, and larger deviations will be given higher weights, and then calculate the average of all residual squares, that is, the Mean Squared Error (MSE). The smaller the value of MSE, the better the prediction performance of the model, because it means that the deviation between the model's predicted value and the actual value is smaller. The Mean Absolute Error (MAE) calculates the average of the absolute values ​​of the difference between the predicted value and the actual value to measure the accuracy of the classification model, which can avoid excessive sensitivity to outliers. The Cross-Entropy Loss function directly reflects the accuracy and confidence of the model's prediction through the difference between the predicted probability distribution and the probability distribution of the true label. By minimizing the cross-entropy loss, the model can learn how to better predict the probability distribution, thereby improving the performance of the classification task.

[0056] In detail, the gradient descent method can be used to optimize the parameters of the classification model until the loss value is less than the preset loss value threshold, wherein the gradient descent method iteratively updates the parameters along the direction of the gradient descent of the loss function until the local minimum is found, and the optimal model parameters are obtained. The optimal model parameters are used to replace the initial model parameters in the classification model to obtain the user classification model.

[0057] Specifically, the gradient descent method includes: Batch Gradient Descent, which means that all data are used to calculate the gradient and update the parameters in each iteration. Stochastic Gradient Descent (SGD), which means that only one sample is used to calculate the gradient and update the parameters in each iteration. Mini-batch Gradient Descent, which means that it is between batch and random, and a small batch of samples is used to calculate the gradient and update the parameters in each iteration. You can also add the momentum of the previous update when updating the parameters, which helps to overcome the local minimum and speed up convergence. Choosing the right gradient descent method and adjusting hyperparameters (such as learning rate, momentum, etc.) are crucial to the performance of the model and can improve the efficiency and accuracy of model training.

[0058] In the embodiment of the present invention, by constructing a user classification model through financial user behavior data, external data resources can be effectively utilized, especially the rich user behavior information contained in the e-commerce platform for model training, thereby improving the richness of user data and obtaining a more accurate user classification model.

[0059] S3. Classify the target financial product users using the user classification model to obtain user categories.

[0060] In the embodiment of the present invention, the target financial product users are a set of users for whom the financial product needs to be promoted. For example, if an insurance company needs to promote pet-related insurance, the target financial product users may be users of smart pet devices on e-commerce platforms, users of Internet insurance services, users with diverse pet types, users on e-commerce platforms who pay high attention to pet health and medical care, and users who keep pets in insurance companies, etc. By classifying each target financial product user, the insurance product purchasing potential of each target financial product user can be identified, and then financial products can be sent in a targeted manner.

[0061] In the embodiment of the present invention, the user classification model is used to classify the target financial product users to obtain user categories, including:

[0062] Acquire target user data of each target financial product user;

[0063] Convert the target user data into a data vector, and use the user classification model to extract features from the data vector to obtain target features;

[0064] The target financial product users are classified according to the target characteristics to obtain user categories.

[0065] In an embodiment of the present invention, the target user data is the user information of the target financial product user, which may be the user's basic information, the user's browsing data and purchase data on the e-commerce platform, as well as the data of each target user in the financial platform, for example, the user's demographic data, the type of financial product purchased, the financial product company purchased, the purchase amount, the purchase time frequency, etc. It may also include the user's past health status, current physical condition, possible health risk factors, and current health insurance status, asset status labels, etc. User classification of target financial product users based on the target user data can help the financial industry better understand and serve existing customers and improve the accuracy of user classification.

[0066] Specifically, the target user data can be first cleaned and converted to remove unnecessary duplicate content, blank lines, etc. to ensure the integrity and correctness of the data. The target user data is then converted into a data vector to facilitate subsequent input into the user classification model to classify the target financial product users. Specifically, text conversion tools such as pre-trained models can be used to convert text into vector data, or word embedding (Word Embedding), Binarizer function can be used for binarization, data can be converted into binary values, and other methods can be used to convert the target user data into a data vector.

[0067] Furthermore, the user classification model can be used to extract features in the data vector to obtain target features. For example, when the user classification model is a logistic regression model, the target features can be constructed by encoding classification variables, and then the output of the linear regression model is mapped to between 0 and 1 using the Sigmoid function (or logistic function) to represent the probability, and the user category of the target user is determined by the probability; when the user classification model is a neural network model, the data vector is feature extracted by the number of layers (depth) of the network and the number of neurons in each layer to obtain the target features, and then the target features are classified using the fully connected layer of the neural network model to obtain the user category of each target financial product user; when the user classification model is a support vector machine, the target features can be classified by finding an optimal hyperplane in the feature space to obtain the user category of each target financial product user.

[0068] In the embodiment of the present invention, by classifying the target financial product users, the purchasing tendency of each target financial product user for the financial product can be obtained, the purchasing potential of each target financial product user can be explored, and then the product can be sent in a targeted manner and the accuracy of the subsequent financial product sending plan can be improved.

[0069] S4. Generate a financial product delivery plan according to the user category, deliver the financial product to the target financial product user according to the financial product delivery plan, and obtain delivery feedback data corresponding to the financial product delivery plan.

[0070] In the embodiment of the present invention, the financial product delivery scheme is a product delivery method for each user, which can recommend the most likely products or services of interest to each target financial product user, so as to improve user experience, increase product transaction rate, and improve user conversion rate.

[0071] In the embodiment of the present invention, the generating of the financial product delivery plan according to the user category includes:

[0072] Calculate the product relevance of each product in a preset product library according to the user category;

[0073] Generate a product delivery list according to the product association;

[0074] A financial product delivery plan is determined according to the product delivery list.

[0075] In an embodiment of the present invention, the product library may be a collection of financial products that need to be promoted, for example, financial products for different objects, financial products with different coverages, financial products with different terms and product types, etc.

[0076] Specifically, the product association degree can be the feature similarity between the user category and each financial product. The product feature vector of each product and the category feature vector of the user category can be extracted, and the feature similarity of the product feature vector and the category feature vector of each user category is calculated. The feature similarity is used as the product association degree, and the product association degree is sorted from large to small to obtain the product sending list of the user category. The product sending list may include the N products with the largest product association degree as the recommended products, and the sorted product sending list may need to be post-processed to meet specific business needs or user experience requirements. For example, products that the user already has can be removed, or it can be ensured that the recommendation list contains products of different categories to increase diversity.

[0077] For example, for target users who have just adopted pets, the most common types of pet insurance can be recommended first, such as pet medical insurance, which can cover the medical expenses required for pets due to illness or accidental injury. This type of insurance usually covers the costs of outpatient, hospitalization, surgery, etc. For pets that need to be transported long distances, pet consignment insurance can be recommended first. This type of insurance is usually purchased during the pet consignment process to protect the pet from possible accidents or death during transportation. If an accident occurs during the consignment process, the insurance company will compensate according to the insurance terms. Therefore, through the user category of each target user, the corresponding financial product can be recommended in real time according to the financial product delivery plan, improving the accuracy of product delivery.

[0078] Among them, cosine similarity and Euclidean distance can be used to measure the product association between product feature vectors and category feature vectors. The calculation formula of cosine similarity is the dot product of two vectors divided by the product of their moduli. The value range of cosine similarity is between -1 and 1. The closer the value is to 1, the more similar the vectors are. Euclidean distance is the simplest and most intuitive distance calculation method. It calculates the straight-line distance between two points in multidimensional space. In the recommendation system, Euclidean distance can be used to calculate the distance between product feature vectors and category feature vectors. The smaller the distance, the higher the feature similarity and the greater the product association.

[0079] In detail, the information, data, and other data reflecting the product attributes of each product in the product delivery list are displayed according to the order of the product delivery list to obtain the financial product delivery plan corresponding to each user category. Among them, a clear and neat layout can be designed in the user interface to ensure that the product pictures, names, prices and key information in the financial product delivery plan are clear at a glance, and a recommendation list in multiple formats is provided, such as list view, grid view, sliding view, etc., to meet the needs of different users; users can be allowed to sort the delivery list according to criteria such as price, rating, new products, sales, etc.; at the same time, the financial product delivery plan can display relevant recommendations based on the user's current context (such as search history, browsing path, seasonal events). For example, when a user browses products in a specific category, other products in the category are recommended.

[0080] In the embodiment of the present invention, sending the financial product to the target financial product user according to the financial product sending scheme includes:

[0081] Generate a sending message according to the financial product sending plan;

[0082] Sending the sending message to a user interface corresponding to the target financial product user;

[0083] The financial product corresponding to the financial product delivery plan is displayed on the user interface.

[0084] Specifically, the sent message contains key information of the financial product in the financial product sending plan, such as name, price, picture, short description, etc. It can be an application sending message, SMS, interface display message, etc. The product is displayed in the user interface corresponding to the target financial product user by sending the message, which can be the homepage carousel, recommendation area, shopping cart page, etc. of the financial platform and e-commerce platform. Ensure that the sent message can be naturally integrated into the user interface and will not interfere with the main browsing experience of the target financial product user.

[0085] In an embodiment of the present invention, a financial product delivery plan can be recommended to target financial product users to obtain direct or indirect reactions from each target financial product user after the financial product is delivered. Product feedback data can be user behavior data such as clicks, browsing, purchases, searches, and collections, user rating data for delivered products, user direct feedback on delivered products, such as "like", "dislike", "not interested", and other data, as well as user interactive behavior data such as sharing, commenting, and liking on social networks. User satisfaction data can also be collected through questionnaires, which can directly help understand the user's overall satisfaction with the delivery strategy. User delivery feedback data can be collected in real time, and the financial product delivery plan can be adjusted in real time to improve the relevance and effect of product delivery.

[0086] S5. Optimize the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan.

[0087] In the embodiment of the present invention, the feedback data can be sent to evaluate the performance of the user classification model, such as precision, recall, F1 score and other indicators, so as to continuously learn and adapt to provide more accurate and personalized recommendations. Precision refers to the proportion of positive categories predicted by the user classification model to those that are actually positive, and recall refers to the proportion of positive categories correctly predicted by the user classification model. F1 score is an indicator for evaluating the performance of the classification model, which combines the balance of precision and recall, and is the harmonic mean of precision and recall.

[0088] In the embodiment of the present invention, the step of optimizing the financial product delivery scheme according to the delivery feedback data to obtain a target delivery scheme includes:

[0089] Utilizing the sent feedback data to optimize the user classification model to obtain an optimized classification model;

[0090] Using the optimized classification model to classify the target financial product users to obtain optimized user categories;

[0091] An optimization recommendation plan is generated according to the user optimization category, and the financial product delivery plan is optimized using the optimization recommendation plan to obtain a target delivery plan.

[0092] In detail, the actual feedback data of each target financial product user in the sent feedback data can be used to optimize the parameters of the user classification model to obtain a more practical optimized classification model. Among them, the model optimization can be the process of training the preset classification model in the above S2, that is, using the sent feedback data as training data, and using the user classification model to classify the sent feedback data to obtain the user optimization category; calculating the loss value of the user classification model according to the user optimization category, and when the loss value is greater than the preset loss value threshold, performing parameter optimization, and returning to the above step of classifying the sent feedback data to obtain the user optimization category; until the loss value is less than the loss value threshold, the optimized classification model is obtained.

[0093] Furthermore, by sending feedback data, the user classification model can be continuously optimized to form a closed-loop iteration, providing more accurate and personalized product recommendations. The optimized sending plan can be used as the target sending plan to achieve more precise product delivery. Among them, the above-mentioned method for generating a financial product sending plan can be used to generate an optimized sending plan, which will not be repeated here.

[0094] Through the implementation of the present invention, the target financial product users can be accurately positioned, and the target financial product users can receive financial product delivery solutions that are more in line with their needs, which can effectively improve the accuracy of product delivery.

[0095] It can be seen that in the above scheme, for complex insurance entities under the insurance business, by screening the financial user behavior data of financial product users, the e-commerce consumption behavior data of financial product users and the insurance product consumption behavior data can be effectively associated; then, the preset classification model can be trained according to the financial user behavior data, and external data resources can be effectively utilized to improve the richness of user data and obtain a more accurate user classification model; the user classification model is then used to classify the target financial product users, generate a financial product sending plan, and obtain corresponding sending feedback data, and the financial product sending plan can be optimized according to the sending feedback data to form a closed-loop iteration, generate a target sending plan that better meets the needs of the target users, and effectively improve the accuracy of financial product recommendations.

[0096] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0097] In one embodiment, a financial product delivery optimization device is provided, and the financial product delivery optimization device corresponds one-to-one to the financial product delivery optimization method in the above embodiment. Figure 4 As shown, the financial product delivery optimization device 400 includes a data screening module 401, a model training module 402, a user classification module 403, a financial product delivery plan generation module 404 and a financial product delivery plan optimization module 405. The functional modules are described in detail as follows:

[0098] The data screening module 401 is used to obtain historical consumption behavior data of non-financial products and screen the financial user behavior data of financial product users according to the historical consumption behavior data;

[0099] A model training module 402 is used to perform model training on a preset classification model using the financial user behavior data to obtain a user classification model;

[0100] A user classification module 403, used to classify target financial product users using the user classification model to obtain user categories;

[0101] A financial product delivery plan generating module 404 is used to generate a financial product delivery plan according to the user category, deliver the financial product to the target financial product user according to the financial product delivery plan, and obtain delivery feedback data corresponding to the financial product delivery plan;

[0102] The financial product delivery plan optimization module 405 is used to optimize the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan.

[0103] In one embodiment, the data screening module 401, when screening the financial user behavior data of the financial product user according to the historical consumption behavior data, is used to:

[0104] De-identifying the historical consumption behavior data to obtain unidentified data;

[0105] Performing joint encryption calculation on the unidentified data to obtain joint user behavior data;

[0106] Financial user behavior data is extracted from the joint user behavior data.

[0107] In one embodiment, the model training module 402, when performing model training on a preset classification model using the financial user behavior data to obtain a user classification model, is used to:

[0108] Using the classification model to classify the financial user behavior data to obtain a predicted category;

[0109] Calculating the loss value of the classification model according to the predicted category, optimizing the parameters of the classification model when the loss value is greater than a preset loss value threshold, and returning to the step of using the classification model to classify the financial user behavior data to obtain the predicted category;

[0110] When the loss value is less than the loss value threshold, a user classification model is obtained.

[0111] In one embodiment, the user classification module 403, when performing user classification on the target financial product user using the user classification model, obtains the user category, which is used for:

[0112] Acquire target user data of each target financial product user;

[0113] Convert the target user data into a data vector, and use the user classification model to extract features from the data vector to obtain target features;

[0114] The target financial product users are classified according to the target characteristics to obtain user categories.

[0115] In one embodiment, the financial product delivery plan generation module 404, when executing the generation of the financial product delivery plan according to the user category, is used to:

[0116] Calculate the product relevance of each product in a preset product library according to the user category;

[0117] Generate a product delivery list according to the product association;

[0118] A financial product delivery plan is determined according to the product delivery list.

[0119] In one embodiment, the financial product delivery plan generation module 404, when executing the delivery of the financial product to the target financial product user according to the financial product delivery plan, is used to:

[0120] Generate a sending message according to the financial product sending plan;

[0121] Sending the sending message to a user interface corresponding to the target financial product user;

[0122] The financial product corresponding to the financial product delivery plan is displayed on the user interface.

[0123] In one embodiment, the financial product delivery plan optimization module 405, when optimizing the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan, is used to:

[0124] Utilizing the sent feedback data to optimize the user classification model to obtain an optimized classification model;

[0125] Using the optimized classification model to classify the target financial product users to obtain optimized user categories;

[0126] An optimized sending plan is generated according to the user optimization category, and the financial product sending plan is optimized using the optimized sending plan to obtain a target sending plan.

[0127] The present invention provides a financial product delivery optimization device, which can effectively associate the e-commerce consumption behavior data of the e-commerce users of the financial product users with the insurance product consumption behavior data by screening the financial user behavior data of the financial product users; then, a preset classification model can be trained according to the financial user behavior data, and external data resources can be effectively used to improve the richness of user data and obtain a more accurate user classification model; the user classification model is then used to classify the target financial product users, generate a financial product delivery plan, and obtain corresponding delivery feedback data, and the financial product delivery plan can be optimized according to the delivery feedback data to form a closed-loop iteration, generate a target delivery plan that better meets the needs of the target users, and effectively improve the accuracy of financial product recommendations.

[0128] For the specific definition of the financial product delivery optimization device, please refer to the definition of the financial product delivery optimization method above, which will not be repeated here. Each module in the above-mentioned financial product delivery optimization device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0129] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a service-side method for optimizing the delivery of financial products.

[0130] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server through a network connection. When the computer program is executed by the processor, it realizes the functions or steps of the client side of a financial product delivery optimization method

[0131] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0132] Acquire historical consumption behavior data of non-financial products, and filter financial user behavior data of financial product users based on the historical consumption behavior data;

[0133] Using the financial user behavior data to perform model training on a preset classification model to obtain a user classification model;

[0134] Using the user classification model to classify target financial product users to obtain user categories;

[0135] generating a financial product delivery plan according to the user category, delivering the financial product to the target financial product user according to the financial product delivery plan, and obtaining delivery feedback data corresponding to the financial product delivery plan;

[0136] The financial product delivery plan is optimized according to the delivery feedback data to obtain a target delivery plan.

[0137] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0138] Acquire historical consumption behavior data of non-financial products, and filter financial user behavior data of financial product users based on the historical consumption behavior data;

[0139] Using the financial user behavior data to perform model training on a preset classification model to obtain a user classification model;

[0140] Using the user classification model to classify target financial product users to obtain user categories;

[0141] generating a financial product delivery plan according to the user category, delivering the financial product to the target financial product user according to the financial product delivery plan, and obtaining delivery feedback data corresponding to the financial product delivery plan;

[0142] The financial product delivery plan is optimized according to the delivery feedback data to obtain a target delivery plan.

[0143] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0144] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0145] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0146] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

[0147] It should be noted that if software tools or components other than those of our company appear in the embodiments of the present application, they are only used for illustrative purposes and do not represent actual use.

Claims

1. A method for optimizing the delivery of financial products, characterized in that: include: Acquire historical consumption behavior data of non-financial products, and filter financial user behavior data of financial product users based on the historical consumption behavior data; Using the financial user behavior data to perform model training on a preset classification model to obtain a user classification model; Using the user classification model to classify target financial product users to obtain user categories; generating a financial product delivery plan according to the user category, delivering the financial product to the target financial product user according to the financial product delivery plan, and obtaining delivery feedback data corresponding to the financial product delivery plan; The financial product delivery plan is optimized according to the delivery feedback data to obtain a target delivery plan.

2. The method for optimizing the delivery of financial products according to claim 1, characterized in that: The step of screening the financial user behavior data of the financial product users according to the historical consumption behavior data includes: De-identifying the historical consumption behavior data to obtain unidentified data; Performing joint encryption calculation on the unidentified data to obtain joint user behavior data; Financial user behavior data is extracted from the joint user behavior data.

3. The method for optimizing the delivery of financial products according to claim 1, characterized in that: The method of using the financial user behavior data to train a preset classification model to obtain a user classification model includes: Using the classification model to classify the financial user behavior data to obtain a predicted category; Calculating the loss value of the classification model according to the predicted category, optimizing the parameters of the classification model when the loss value is greater than a preset loss value threshold, and returning to the step of using the classification model to classify the financial user behavior data to obtain the predicted category; When the loss value is less than the loss value threshold, a user classification model is obtained.

4. The method for optimizing the delivery of financial products according to claim 1, characterized in that: The user classification model is used to classify the target financial product users to obtain user categories, including: Acquire target user data of each target financial product user; Convert the target user data into a data vector, and use the user classification model to extract features from the data vector to obtain target features; The target financial product users are classified according to the target characteristics to obtain user categories.

5. The method for optimizing the delivery of financial products according to claim 1, characterized in that: Generating a financial product delivery plan according to the user category includes: Calculate the product relevance of each product in a preset product library according to the user category; Generate a product delivery list according to the product association; A financial product delivery plan is determined according to the product delivery list.

6. The method for optimizing the delivery of financial products according to claim 1, characterized in that: Sending the financial product to the target financial product user according to the financial product sending plan includes: Generate a sending message according to the financial product sending plan; Sending the sending message to a user interface corresponding to the target financial product user; The financial product corresponding to the financial product delivery plan is displayed on the user interface.

7. The method for optimizing the delivery of financial products according to claim 1, characterized in that: The step of optimizing the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan includes: Utilizing the sent feedback data to optimize the user classification model to obtain an optimized classification model; Using the optimized classification model to classify the target financial product users to obtain optimized user categories; An optimized sending plan is generated according to the user optimization category, and the financial product sending plan is optimized using the optimized sending plan to obtain a target sending plan.

8. A financial product delivery optimization device, characterized in that: include: A data screening module, used to obtain historical consumption behavior data of non-financial products, and screen financial user behavior data of financial product users based on the historical consumption behavior data; A model training module, used to perform model training on a preset classification model using the financial user behavior data to obtain a user classification model; A user classification module, used to classify target financial product users using the user classification model to obtain user categories; A financial product delivery plan generating module, configured to generate a financial product delivery plan according to the user category, deliver the financial product to the target financial product user according to the financial product delivery plan, and obtain delivery feedback data corresponding to the financial product delivery plan; The financial product delivery plan optimization module is used to optimize the financial product delivery plan according to the delivery feedback data to obtain a target delivery plan.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the financial product delivery optimization method according to any one of claims 1 to 7 are implemented.

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

Images