Credit card application processing method and device, equipment, medium and product

Abstract

The invention provides a credit card application processing method and device, equipment, a medium and a product, relates to the field of artificial intelligence, and can be applied to the field of financial science and technology. Comprising the steps that credit card information of a target user is acquired, and the target user is a user who fills in credit card application information but does not submit the credit card application information; the credit card information of the target user is analyzed by adopting a caching technology and a data matching algorithm, intra-bank credit card information of the target user is determined, and the intra-bank credit card information is used for indicating that an intra-bank credit card is held or not held; if the intra-bank credit card information represents that the target user does not hold the intra-bank credit card, target feature information in the credit card information is extracted according to an auto-encoder algorithm; according to the target feature information and the matrix decomposition algorithm, the first service information is pushed to the target user, the first service information is information of at least one target credit card determined according to the demand of the target user, the application efficiency of the credit card is improved, and the satisfaction degree of credit card application of the user is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, and in particular to a credit card application processing method and device, equipment, medium and product. BACKGROUND

[0002] With the continuous development of Internet technology, the traditional financial industry is facing a huge change, which not only changes the way of providing financial services, but also deeply affects the structure of the financial industry. Under this background, many financial services including credit card applications begin to migrate to online channels.

[0003] The online credit card application system can automatically process the information submitted by the user and give the audit result in a short time. Compared with the traditional manual audit method, online application is more efficient and faster.

[0004] However, the online application will appear the phenomenon that the user exits the application process after submitting personal information, which causes resource waste and reduces the operation efficiency of the credit card system. SUMMARY

[0005] The present application provides a credit card application processing method, device, equipment, medium and product to solve the technical problem that the online credit card application appears the phenomenon that the user exits the application process after submitting personal information, which causes resource waste and reduces the operation efficiency of the credit card system.

[0006] In a first aspect, the present application provides a credit card application processing method, comprising:

[0007] Obtaining credit card information of a target user, the target user being a user who fills in the credit card application information but does not submit, and the credit card information including personal information, contact information and financial status of the target user;

[0008] Using a cache technology and a data matching algorithm to analyze the credit card information of the target user, and determining the in-house credit card information of the target user, the in-house credit card information being used to indicate whether the target user holds an in-house credit card or not;

[0009] If the in-house credit card information represents that the target user does not hold an in-house credit card, extracting target feature information in the credit card information according to a self-encoder algorithm, the target feature information including consumption characteristics and needs of the target user;

[0010] According to the target feature information and a matrix decomposition algorithm, pushing first service information to the target user, the first service information being information of at least one target credit card determined according to the needs of the target user.

[0011] In a second aspect, the application provides a credit card application processing device, comprising:

[0012] a credit card information obtaining module, configured to obtain credit card information of a target user, the target user being a user who has filled in credit card application information but has not submitted, and the credit card information comprising personal information, contact information and financial status of the target user;

[0013] an in-house credit card information determining module, configured to analyze the credit card information of the target user by using a caching technology and a data matching algorithm, and determine in-house credit card information of the target user, the in-house credit card information being used to indicate whether the target user holds an in-house credit card or not;

[0014] a target feature information extracting module, configured to, if the in-house credit card information indicates that the target user does not hold an in-house credit card, extract target feature information in the credit card information according to a self-encoder algorithm, the target feature information comprising consumption features and demands of the target user;

[0015] a first service information pushing module, configured to push first service information to the target user according to the target feature information and a matrix decomposition algorithm, the first service information being information of at least one target credit card determined according to the demands of the target user.

[0016] In a third aspect, the application provides an electronic device, comprising:

[0017] a processor, and a memory connected to the processor in communication;

[0018] the memory stores computer execution instructions;

[0019] the processor executes the computer execution instructions stored in the memory, so as to implement the method according to the first aspect.

[0020] In a fourth aspect, the application provides a computer readable storage medium, the computer readable storage medium storing computer execution instructions, the computer execution instructions being executed by a processor to implement the method according to the first aspect.

[0021] In a fifth aspect, the application provides a computer program product, comprising a computer program, the computer program being executed by a processor to implement the method according to the first aspect.

[0022] The credit card application processing method, device, equipment, medium and product provided by the application, by obtaining the credit card information of the target user, the target user being a user who fills in the credit card application information but does not submit, the credit card information including the personal information, contact information and financial status of the target user; the credit card information of the target user is analyzed by using the cache technology and the data matching algorithm, and the in-house credit card information of the target user is determined, the in-house credit card information being used to indicate whether the target user holds an in-house credit card or not; if the in-house credit card information indicates that the target user does not hold an in-house credit card, the target feature information in the credit card information is extracted according to the self-encoder algorithm, the target feature information including the consumption feature and demand of the target user; according to the target feature information and the matrix decomposition algorithm, the target user is pushed the first service information, the first service information being the information of at least one target credit card determined according to the demand of the target user, the credit card application efficiency is improved, and the satisfaction of the user with the credit card application is improved. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are incorporated herein and form part of the specification, illustrate embodiments consistent with the present application and, together with the description, further serve to explain the principles of the application.

[0024] Figure 1 The overall architecture schematic diagram of the credit card application processing method provided by the application is shown in the figure.

[0025] Figure 2 The flowchart of the first embodiment of the credit card application processing method provided by the application is shown in the figure.

[0026] Figure 3 The flowchart of the second embodiment of the credit card application processing method provided by the application is shown in the figure.

[0027] Figure 4 The flowchart of the third embodiment of the credit card application processing method provided by the application is shown in the figure.

[0028] Figure 5 The flowchart of the fourth embodiment of the credit card application processing method provided by the application is shown in the figure.

[0029] Figure 6 The flowchart of the fifth embodiment of the credit card application processing method provided by the application is shown in the figure.

[0030] Figure 7 The flowchart of the sixth embodiment of the credit card application processing method provided by the application is shown in the figure.

[0031] Figure 8 The structure schematic diagram of the credit card application processing device provided by the application is shown in the figure.

[0032] Figure 9 The structure schematic diagram of the electronic device provided by the application is shown in the figure.

[0033] The specific embodiments of the application will be described in detail below with reference to the accompanying drawings. These drawings and the following description are not intended to limit the scope of the application in any way, but are merely intended to illustrate the concepts underlying the application. DETAILED DESCRIPTION

[0034] The exemplary embodiments will be described in detail below with reference to the accompanying drawings. The following description is not intended to limit the scope of the application in any way, but is merely intended to illustrate the concepts underlying the application.

[0035] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards of relevant countries and regions, necessary security measures are taken, public order and good customs are not violated, and appropriate operation portals are provided for users to choose authorization or refusal.

[0036] And the present application involves big data analysis of user information (including but not limited to personal biological characteristics, identity data, consumption data, asset data, electronic terminal operation data, etc.), and uses artificial intelligence technology for automatic decision-making, and makes technical solutions based on automatic decision-making results that have a significant impact on personal rights and interests, provides appropriate operation portals for users to choose to agree or refuse automatic decision-making results; if the user chooses to refuse, the expert decision-making process is entered.

[0037] At present, there are the following problems in the analysis and recommendation of user credit card information: first, there is a lack of efficient means to quickly determine whether the user already holds an in-house credit card, leading to repeated analysis and resource waste; second, existing credit card applications often fail to deeply mine the potential needs of users, and the matching of credit card features and user features lacks precision, making it difficult to provide personalized services that truly meet the needs of users; third, the processing of user interaction behavior in the recommendation process is limited, and these problems affect the quality of bank credit card recommendation services and user experience.

[0038] To solve the above problems, the technical scheme combines cache technology and data matching algorithm to realize fast judgment and accurate analysis of user credit card information, avoid repeated calculation, further uses the self-encoder algorithm to extract user potential demand, decomposes and optimizes user and credit card features through matrix decomposition algorithm, accurately matches user demand, finally combines the optimization learning algorithm and gradient descent method to continuously optimize the recommendation result, and generates targeted credit card recommendation service. Through this scheme, the problems of low efficiency and insufficient personalization in traditional technology are solved, an efficient and accurate credit card recommendation system is provided for users, and the service ability and user satisfaction of the bank are improved.

[0039] It should be noted that the credit card application processing method, device, equipment, medium and product provided by the present application can be used in the field of financial technology, and can also be used in any field other than the field of financial technology. The application field of the credit card application processing method, device, equipment, storage medium and product in the present application is not limited.

[0040] Figure 1 The overall architecture of the credit card application processing method provided by the present application is shown in the figure. As shown in the figure, Figure 1 The overall architecture 10 of the credit card application processing method includes: a user information collection module 101 responsible for collecting credit card application information submitted by a target user, including personal information, contact information and financial status. A cache and data matching module 102: using cache technology to quickly determine whether the user already holds an in-house credit card; if the cache is not hit, the user information is compared with the in-house credit card database through the data matching algorithm to determine the user's credit card holding status, and the judgment efficiency is improved. A feature extraction module 103: based on the self-encoder algorithm, the target user information is preprocessed and analyzed, and the target feature information reflecting the user's consumption behavior, demand preference, etc. is extracted, providing core data support for the recommendation system. A recommendation engine module 104: an interaction matrix of user-credit card items is constructed through a matrix decomposition algorithm, the user feature and credit card feature matrices are decomposed and optimized, the interest degree of the user to the credit card product is accurately predicted, and a recommended credit card list is generated. A service pushing module 105: according to the output of the recommendation engine, the first service information (such as target credit card recommendation) or the second service information (such as adjustable benefits of existing credit cards) is pushed to the user, and personalized pushing is realized through various channels (such as short message, application program). An optimization and feedback module 106: the recommendation model is continuously optimized by using the gradient descent method, and feedback data is collected through the user's interaction behavior (such as clicking or application result), and the recommendation accuracy and user experience are continuously improved.

[0041] The overall architecture 10 of the credit card application processing method takes user demand as the core, realizes the full-process automation of information collection, analysis, recommendation and pushing through modular design, and significantly improves the processing efficiency and service quality of credit card application.

[0042] The technical solutions of the present application and how the technical solutions solve the above technical problems will be described in detail below with specific examples. The following specific examples can be combined with each other, and the same or similar concepts or processes can not be described again in some examples. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0043] Figure 2 The flowchart of the credit card application processing method embodiment provided by the present application is shown in Figure Figure 1 The method comprises:

[0044] S201, obtaining the credit card information of a target user.

[0045] The target user refers to a user who fills in the credit card application information but does not submit. These target users show potential demand for credit cards. The credit card information includes the personal information of the target user (such as name, age), contact information (such as phone number, email) and financial status (such as income level, asset information).

[0046] Further, the credit card information of the target user is stored on multiple nodes of the blockchain.

[0047] The blockchain is a distributed ledger technology that stores data on multiple nodes, has the characteristics of non-tamperability, high security and traceability. In addition, each node on the blockchain coordinates work through a consensus algorithm, ensuring data synchronization and consistency.

[0048] The multiple nodes refer to the distributed storage units in the blockchain, and each node saves complete or partial ledger data. In the embodiments of the present application, these nodes can be multiple servers, data centers or partner nodes of the bank.

[0049] Specifically, the credit card information of the target user can be encrypted using an encryption algorithm to protect the security of the private data. The encrypted data can only be decrypted when authorized by the node or the user. Then the encrypted credit card information is sent to each node in the blockchain, and each node verifies the legality of the credit card information through a consensus mechanism. When the verification is passed, the credit card information is recorded in the blockchain ledger of each node. In addition, all nodes will store the same data copy to ensure data consistency. Through the blockchain explanation, the security, non-tamperability and storage stability of the credit card information can be improved, ensuring that the user's credit card information can be efficiently and reliably managed.

[0050] S202, using a caching technology and a data matching algorithm to analyze the credit card information of the target user, and determining the in-line credit card information of the target user.

[0051] In the embodiments of the present application, the cache technology is used to temporarily store the user information and the result of the previous comparison, so as to avoid repeated comparison and improve the information processing efficiency.

[0052] The data matching algorithm is used to compare the input data with the existing data according to certain rules or methods, for verification, query or association matching. In the embodiments of the present application, the data matching algorithm is used to convert the user information into a hash value by using a hash function, and compare the hash value with the hash index of the in-house credit card database, so as to determine whether the user holds an in-house credit card. The in-house credit card information is used to indicate whether the user holds an in-house credit card or not.

[0053] Specifically, first, the cache technology is used to search the credit card record of the target user in the cache. If the cache hits, it is directly returned that the target user holds an in-house credit card. If the cache misses, the credit card information of the user is processed by the hash function according to the data matching algorithm, a unique hash value is generated, and the hash value is compared with the hash index in the in-house credit card database. If the hash values match, the flag of "the user holds an in-house credit card" is returned. If the flag of "the user does not hold an in-house credit card" is returned, the credit card information of the target user is continuously processed.

[0054] S203, if the in-house credit card information indicates that the target user does not hold an in-house credit card, the target feature information in the credit card information is extracted according to the auto-encoder algorithm.

[0055] The auto-encoder algorithm is an unsupervised learning model used to extract low-dimensional key features from high-dimensional data. The target feature information refers to the core features of the credit card information of the target user that are meaningful for recommending credit cards to the user, such as the consumption features and needs of the target user.

[0056] Specifically, the credit card information filled by the target user (such as income, age, and consumption preference) is used as the input of the auto-encoder, and then the encoder extracts the key features (such as the preference of the target user for benefits and the demand for credit limit) in the credit card information. The decoder reduces the noise by learning to extract the key data. Finally, the consumption features (such as shopping preference) and demand features (such as credit limit preference or fee-free demand) of the target user are extracted from the low-dimensional features output by the auto-encoder model.

[0057] S204, according to the target feature information and the matrix decomposition algorithm, the first service information is pushed to the target user.

[0058] The matrix decomposition algorithm is a classical mathematical and machine learning method for extracting low-dimensional latent features from high-dimensional data to solve data coefficient problems and achieve prediction. In the embodiments of the present application, the matrix decomposition algorithm is used to decompose the interaction matrix between users and credit card products into a user feature matrix and a product feature matrix to predict the interest of the target user in uncontacted products (such as credit card types).

[0059] The first service information refers to information of at least one target credit card determined according to the needs of the target user, such as a specific credit card type or a preferential activity.

[0060] Specifically, an interaction matrix between users and credit cards can be constructed according to the historical behavior data (such as click or application records) of the target user, and then the matrix decomposition algorithm is used to decompose the interaction matrix between users and credit cards into feature information of the target user and match it with the latent preference in the credit card matrix, thereby predicting the interest of the user in uncontacted credit cards. Finally, according to the interest ranking, the target user is pushed the first service information (such as recommended cashback cards or travel benefit cards).

[0061] The credit card application processing method provided in the embodiments collects the credit card information of the target user, analyzes the data, and recommends and pushes the process, efficiently identifies the needs of the user and accurately pushes the service information, improves the user experience, and provides strong technical support for the bank to optimize the application process of the credit card.

[0062] Figure 3 The flowchart of the credit card application processing method provided in Embodiment Two of the present application is shown in FIG. 2. Figure 3 As shown in FIG. 2, based on the embodiments, the method further includes: Figure 2

[0063] S301, if the in-house credit card information of the target user indicates that the target user holds an in-house credit card, the feature information of the in-house credit card held by the target user is obtained.

[0064] The feature information of the in-house credit card includes the credit card type, the credit card limit, the usage, the benefit enjoyment, and the annual fee.

[0065] Specifically, if the in-house credit card information of the target user indicates that the target user already holds an in-house credit card, the detailed information of the credit card held by the target user can be extracted from the internal system of the bank, and then the data matching technology is used to quickly locate the credit card record of the target user. At the same time, the credit card information of the target user is stored in a structured manner to ensure that each feature information is complete and accurate for subsequent satisfaction analysis of the credit card. ​

[0066] S302, determining the satisfaction information of the target user to the in-circle credit card according to the characteristic information of the in-circle credit card held by the target user and a logistic regression algorithm.

[0067] The logistic regression algorithm is a classification model for predicting the probability of a target variable, and the satisfaction information reflects the subjective feelings of the target user to the attribute information of the credit card held by the target user, including credit limit satisfaction, benefit enjoyment satisfaction, and annual fee satisfaction.

[0068] Specifically, the characteristic information extracted in S301 is taken as the model input of the logistic regression algorithm, and the model is trained. The model formula is:

[0069]

[0070] Wherein, x represents the input characteristics, such as credit limit, annual fee, etc. w, b represent model parameters. y represents the prediction result, such as satisfied or not satisfied.

[0071] If the input characteristic information is the credit limit, and the final credit limit satisfaction score is 80%, it means that the credit card held by the target user may need to improve the credit limit. If the input characteristic information is the benefit enjoyment situation, and the final credit limit satisfaction score is 60%, it means that the credit card held by the target user may need to increase the benefit.

[0072] S303, pushing second service information to the target user according to the satisfaction information.

[0073] The second service information is a personalized adjustment service of the credit card provided for the target user according to the satisfaction analysis result, including information of at least one adjustable service of the in-circle credit card of the target user. For example, if the target user is not satisfied with the credit card, the target user is pushed with the service of "applying for increasing the credit limit of the credit card"; if the target user is not satisfied with the benefit of the credit card, the target user is pushed with the service of "point exchange instruction"; if the target user is not satisfied with the annual fee of the credit card, the target user is pushed with the service of "first year free annual fee upgrade card" or "consumption full reduction annual fee". The above push content can be pushed to the target user through short message, application program, email, etc. After receiving the second service information, the target user can operate one key to directly apply for the related service or adjustment.

[0074] The credit card application processing method provided by the embodiment of the application lays a data foundation for satisfaction analysis by acquiring characteristic information of a credit card held by a target user. Then, the satisfaction of the target user to the credit card characteristics (limit, benefits, annual fee, etc.) is analyzed by using a logistic regression algorithm. Finally, personalized credit card adjustment services are pushed to the target user according to the satisfaction result, which effectively improves the accuracy of bank services and the satisfaction of the target user, and promotes the continuous use of credit card services by the target user.

[0075] Figure 4 The flowchart of the third embodiment of the credit card application processing method provided by the application is shown in FIG. 3. Figure 4 As shown in FIG. 3, on the basis of the first embodiment, Figure 2 the credit card information of the target user is analyzed by using a cache technology and a data matching algorithm to determine the in-house credit card information of the target user, including:

[0076] S401, analyzing the credit card information of the target user according to the cache technology to determine whether the target user is a user with an identifier.

[0077] The identifier is the identifier of the user who has held an in-house credit card. Specifically, when the target user submits the credit card information, the identifier of the target user is first searched in the cache, for example, by matching the record in the cache through the identity information (such as the ID number or mobile phone number) of the user.

[0078] S402, if the target user is a user with an identifier, the in-house credit card information of the target user is determined to be holding an in-house credit card.

[0079] In this step, if the information of the target user exists in the cache and has an identifier, it indicates that the target user has held an in-house credit card, so further calculation or matching is not needed, and the credit card information of the target user can be directly used as the input information of the subsequent process.

[0080] S403, if the target user is a user without an identifier, the credit card information of the target user is analyzed according to the data matching algorithm to determine the in-house credit card information of the target user.

[0081] In this step, if there is no identifier of the target user in the cache, it indicates that the target user does not hold an in-house credit card. At this time, the credit card information of the target user needs to be hashed, and the record matching the hash value of the target user is searched in the in-house credit card database. If the matching hash value is found in the in-house credit card database, it indicates that the target user has held an in-house credit card.

[0082] The credit card application processing method provided in the embodiments of the present application uses a cache technology to quickly find the credit card information of a target user, and if the cache is hit, it is directly determined that the target user holds an in-house credit card. For a user who has been confirmed in the cache, it is directly determined that the user holds an in-house credit card, and no further processing is required. For a user who does not hit the cache, a data matching algorithm is used to compare with a database to determine whether the user holds an in-house credit card. Through the combination of the cache technology and the data matching algorithm, the accuracy of the matching result can be determined.

[0083] Figure 5 A flowchart of the credit card application processing method provided in Embodiment Four of the present application is shown in FIG. 4. Figure 5 As shown in FIG. 4, based on Embodiment One, Figure 4 if the target user is an unidentified user, the credit card information of the target user is analyzed according to a data matching algorithm to determine the in-house credit card information of the target user, including:

[0084] S501, hash processing is performed on the identity information of each user in the database to generate a hash value corresponding to the identity information of each user.

[0085] The hash processing refers to using a hash algorithm to convert the identity information of the target user into a unique fixed-length hash value. It should be noted that the hash value is irreversible, that is, the original data cannot be derived from the generated hash value, which can ensure the security of the information of the target user. The identity information of all users who have held an in-house credit card is stored in the database, and each hash value corresponds to each user one by one;

[0086] Specifically, the database is traversed, and the identity information of each user who holds an in-house credit card in the database is read one by one. Then, a specified hash algorithm is used to convert the identity information of each user into a hash value, and the generated hash value is one-to-one corresponding to the user identity information to form a new database index for subsequent quick comparison.

[0087] S502, according to the credit card information of the target user, a to-be-compared hash value of the target user is generated.

[0088] The to-be-compared hash value refers to the identity information in the credit card information submitted by the target user is subjected to hash processing to generate a unique hash value. Specifically, the identity information (such as an ID card number, a mobile phone number) in the credit card information of the target user is extracted, and then the same hash algorithm as that of the database is used to generate the to-be-compared hash value.

[0089] S503, the hash value corresponding to the identity information of each user is compared with the to-be-compared hash value of the target user to obtain a comparison result.

[0090] Wherein, the comparison result is used to represent whether the to-be-compared hash value of the target user matches the generated hash value in the database. Specifically, by using the hash index, the record consistent with the to-be-compared hash value of the target user is directly found from the database. Then the comparison result is generated, if the comparison result represents that there is a matching value, the information of "matching success" is returned, if the comparison result represents that there is no matching value, the information of "unmatching success" is returned.

[0091] S504, if the comparison result represents that there is a hash value in the database matching the to-be-compared hash value of the target user, it is determined that the target user has the in-house credit card information.

[0092] In this step, if it is determined according to the comparison result that the target user already exists in the in-house credit card database, the target user is marked as "having an in-house credit card", and this information is taken as the input of the subsequent process (such as recommending a credit card or analyzing satisfaction).

[0093] S505, if the comparison result represents that there is no hash value in the database matching the to-be-compared hash value of the target user, it is determined that the target user does not have the in-house credit card information.

[0094] In this step, if it is determined according to the comparison result that the target user does not exist in the in-house credit card database, the target user is marked as "not having an in-house credit card", and this information is taken as the input of the subsequent process (such as recommending a new card or preferential service).

[0095] The credit card application processing method provided by the embodiment of the application constructs a fast and efficient user identity comparison process, realizes accurate matching of user credit card information through a data matching algorithm and a hash algorithm, and can efficiently and accurately determine whether the target user has an in-house credit card, thereby providing a reliable basis for subsequent recommendation or analysis.

[0096] Figure 6 The flowchart of the credit card application processing method provided by the fifth embodiment of the application is shown in FIG. 6. Figure 6 As shown in FIG. 6, on the basis of the embodiment, Figure 2 if the in-house credit card information represents that the target user does not have an in-house credit card, the target feature information in the credit card information of the target user is extracted according to the self-encoder algorithm, including:

[0097] S601, if the target user does not have an in-house credit card, the credit card information of the target user is preprocessed to obtain the feature information of the target user.

[0098] Wherein, the preprocessing refers to converting the credit card information of the target user into a structured and standardized data line, removing redundant and irrelevant information, and ensuring the accuracy of subsequent analysis.

[0099] Specifically, first, remove the target user's credit card information of the repetition, missing or error data, such as, clean outliers (such as negative income fill). Then the credit card information is normalized or standardized to ensure data distribution consistency, such as, the income is expressed in the same unit (such as monthly income), and the consumption amount is normalized to [0, 1]. Finally, more potential useful features are generated from the above processed information, such as, consumption frequency, total income ratio, different industry classification of occupation.

[0100] S602, according to the autoencoder algorithm, construct an autoencoder model.

[0101] Wherein, the autoencoder algorithm is an unsupervised learning model, commonly used for feature extraction and dimensionality reduction, and the autoencoder consists of two parts, one part is the encoder, used to compress high-dimensional input data into low-dimensional latent feature identification, the other part is the decoder, used to restore the low-dimensional representation to the original input data. The autoencoder model is a model structure constructed according to the autoencoder algorithm, mainly including input layer, hidden layer (latent feature representation layer) and output layer.

[0102] Specifically, the feature information of the target user (such as a multi-dimensional numerical vector) is taken as the input layer of the autoencoder model, and the hidden layer of the autoencoder model is taken as the intermediate layer to extract low-dimensional target feature information, and the output layer of the autoencoder model is taken as the reconstructed input feature to supervise the training. Then initialize the autoencoder model, set the autoencoder model parameters, such as the dimension of the hidden layer, the activation function and the learning rate, etc., and then optimize the target of the autoencoder model, that is, the loss function is used to minimize the error between the input feature and the decoded reconstructed feature.

[0103] S603, according to the feature information of the target user, train the autoencoder model to extract the target feature information in the feature information of the target user.

[0104] Wherein, the training refers to using the feature information of the target user as input to learn the potential target feature representation through the autoencoder model. The target feature information refers to the key features extracted from the credit card information of the target user, which can directly reflect the demand or preference of the target user, such as, consumption category (travel or shopping), interest in free annual fee or high cashback.

[0105] Specifically, the preprocessed feature information of S601 is taken as the input of the autoencoder, and then the autoencoder model is optimized by gradient descent method, and then the potential features of the input data are learned. Finally, using the trained encoder part, the high-dimensional features of the target user are compressed into low-dimensional latent representation, wherein the target feature information examples are user consumption frequency, high priority benefit demand.

[0106] The credit card application processing method provided by the embodiment of the application forms structured feature information by cleaning, standardizing and feature generating the credit card information of the target user; then constructs a model based on a self-encoder algorithm, designs a neural network structure for extracting latent features; and finally extracts core features (target feature information) of the target user by training the self-encoder model, accurately identifies the demand of the target user, and lays a foundation for subsequent personalized recommendation.

[0107] Figure 7 The flowchart of the credit card application processing method provided by the sixth embodiment of the application is shown in FIG. 6. Figure 7 As shown in FIG. 6, Figure 2 based on the embodiment, the first service information is pushed to the target user according to the target feature information and the matrix decomposition algorithm, including:

[0108] S701, constructing a user-item interaction matrix according to the target feature information and the credit card item information.

[0109] The credit card information represents the credit card products provided by the bank and related information thereof, including the credit card product types such as shopping cards, cashback cards and air mileage cards.

[0110] The row in the user-item interaction matrix represents the target user, and the column represents the credit card item. Each element in the matrix represents the interaction degree of the target user with the credit card item.

[0111] Specifically, the feature information of the target user is matched with the credit card item information to obtain a preliminary interaction matrix. For example, if the target user prefers a high cashback credit card, the interaction score of this type of card is higher. Then the interaction matrix is constructed.

[0112] S702, decomposing and processing the target user-item interaction matrix according to the matrix decomposition algorithm to obtain a user feature matrix and an item feature matrix.

[0113] The matrix decomposition algorithm is to decompose the user-item interaction matrix into two low-dimensional matrices, including a user feature matrix and an item feature matrix. The user feature matrix is used to represent the potential demand and preference of the target user, and the item feature matrix is used to represent the attributes and characteristics of the credit card item.

[0114] The formula can be expressed as: R≈UV. Wherein, R represents the user-item interaction matrix, U represents the user feature matrix, and V represents the item feature matrix.

[0115] Specifically, the user-item interaction matrix R is taken as input, and R is decomposed into U and V using a matrix decomposition algorithm, U represents a feature vector corresponding to each row of a user, and V represents a feature vector corresponding to each column of a credit card item. Finally, the user feature matrix U is output to extract the potential demand of the target user, and the item feature matrix V is output to extract the attributes of the credit card item.

[0116] S703, by optimizing the learning algorithm and the gradient descent method, the user feature matrix and the item feature matrix are optimized, and the demand of the target user for the credit card is determined.

[0117] Among them, the optimization learning algorithm is to adjust the decomposed user feature matrix and item feature matrix to improve the prediction accuracy. Gradient descent method is an optimization algorithm, which can minimize the error between actual interaction data and predicted value by continuously adjusting matrix parameters.

[0118] S704, according to the demand of the target user for the credit card, the information of at least one target credit card is determined.

[0119] Among them, the information of the target credit card refers to the credit card product information highly matched with the demand of the target user, including product name, characteristics, benefit introduction, etc.

[0120] Specifically, according to the user feature matrix U and the item feature matrix V, the interest degree of the target user for the non-interaction credit card is calculated, that is, the interest degree is U i V j . Then, according to the interest degree, the credit card with the highest score is screened out, for example, if the user prefers shopping cashback, the credit card with high cashback ratio is recommended, and if the user pays attention to travel discount, the airline mileage card or hotel benefit card is recommended.

[0121] S705, according to the information of at least one target credit card, the first service information is pushed to the target user.

[0122] In this step, personalized recommendation content can be generated according to the demand of the target user and the characteristics of the target credit card, for example, based on your consumption habits, a shopping credit card is recommended, which can enjoy 3% shopping cashback and free service.

[0123] The credit card application processing method provided by the embodiment of the application can accurately match the demand of the user by constructing a user-item interaction matrix, quantifying the relationship between the user and the credit card item, then extracting the potential features of the user and the item through matrix decomposition, optimizing the feature matrix, accurately determining the specific demand of the user for the credit card, finally screening out the target credit card highly related to the demand of the user, and pushing personalized recommendation information, so as to realize personalized recommendation of credit card products.

[0124] Figure 8A structure schematic diagram of a credit card application processing device provided in the present application is shown in FIG. 8. As shown in FIG. 8, the credit card application processing device 80 comprises: Figure 8

[0125] a credit card information obtaining module 801, configured to obtain credit card information of a target user, the target user being a user who has filled in credit card application information but has not submitted, and the credit card information comprising personal information, contact information and financial information of the target user.

[0126] an in-house credit card information determining module 802, configured to analyze the credit card information of the target user by using a caching technique and a data matching algorithm, and determine in-house credit card information of the target user, the in-house credit card information being used to indicate whether the target user holds an in-house credit card or not.

[0127] a target feature information extracting module 803, configured to, if the in-house credit card information indicates that the target user does not hold an in-house credit card, extract target feature information in the credit card information according to a self-encoder algorithm, the target feature information comprising consumption features and demands of the target user.

[0128] a first service information pushing module 804, configured to push first service information to the target user according to the target feature information and a matrix decomposition algorithm, the first service information being information of at least one target credit card determined according to the demands of the target user.

[0129] Further, the credit card information obtaining module 801 is further specifically configured to:

[0130] if the in-house credit card information indicates that the target user holds an in-house credit card, obtain feature information of the in-house credit card held by the target user;

[0131] determine satisfaction information of the target user with the in-house credit card according to the feature information of the in-house credit card held by the target user and a logistic regression algorithm;

[0132] push second service information to the target user according to the satisfaction information, the second service information comprising information of at least one adjustable service of the in-house credit card of the target user;

[0133] wherein the feature information of the in-house credit card comprises a credit card type, a credit card limit, a usage, a benefit enjoyment and an annual fee;

[0134] the satisfaction information comprises a credit card limit satisfaction, a benefit enjoyment satisfaction and an annual fee satisfaction.

[0135] Further, the in-house credit card information determining module 802 is further specifically configured to:

[0136] ​According to the cache technology, the credit card information of the target user is analyzed to determine whether the target user is a user with an identity, which is an identity of a user who has held an in-house credit card;

[0137] If the target user is a user with an identity, it is determined that the in-house credit card information of the target user is holding an in-house credit card.

[0138] If the target user is a user without an identity, the credit card information of the target user is analyzed according to a data matching algorithm to determine the in-house credit card information of the target user.

[0139] Further, the in-house credit card information determination module 802 is further specifically used for:

[0140] If the target user is a user without an identity, the credit card information of the target user is analyzed according to a data matching algorithm to determine the in-house credit card information of the target user, comprising:

[0141] Hash processing is performed on the identity information of each user in the database to generate a hash value corresponding to the identity information of each user, wherein the identity information of each user holding an in-house credit card is stored in the database, and each hash value corresponds to each user one by one.

[0142] According to the credit card information of the target user, a to-be-compared hash value of the target user is generated.

[0143] The hash value corresponding to the identity information of each user is compared with the to-be-compared hash value of the target user to obtain a comparison result.

[0144] If the comparison result represents that there is a hash value in the database that matches the to-be-compared hash value of the target user, it is determined that the in-house credit card information of the target user is holding an in-house credit card.

[0145] If the comparison result represents that there is no hash value in the database that matches the to-be-compared hash value of the target user, it is determined that the in-house credit card information of the target user is not holding an in-house credit card.

[0146] Further, the target feature information extraction module 803 is further specifically used for:

[0147] If the target user does not hold an in-house credit card, the credit card information of the target user is preprocessed to obtain feature information of the target user.

[0148] According to the auto-encoder algorithm, an auto-encoder model is constructed.

[0149] According to the feature information of the target user, the auto-encoder model is trained to extract target feature information from the feature information of the target user.

[0150] Further, the first service information pushing module 804 is further specifically used for:

[0151] According to the target feature information and the credit card item information, a user-item interaction matrix is constructed, wherein the credit card item information comprises a credit card product type;

[0152] According to a matrix decomposition algorithm, the target user-item interaction matrix is decomposed to obtain a user feature matrix and an item feature matrix;

[0153] By using an optimization learning algorithm and a gradient descent method, the user feature matrix and the item feature matrix are optimized to determine the demand of the target user for the credit card;

[0154] According to the demand of the target user for the credit card, information of at least one target credit card is determined;

[0155] According to the information of the at least one target credit card, the first service information is pushed to the target user.

[0156] Further, the credit card application processing apparatus 80 is further specifically used for:

[0157] The credit card information of the target user is stored on the plurality of nodes of the block chain.

[0158] Figure 9 A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 9. Figure 9 As shown in FIG. 9, the electronic device 90 comprises:

[0159] The electronic device 90 can comprise a processor 901 with one or more processing cores, a memory 902 with one or more computer readable storage media, a communication component 903, and the like. The processor 901, the memory 902, and the communication component 903 are connected through a bus 904.

[0160] In the specific implementation process, the at least one processor 901 executes the computer execution instructions stored in the memory 902, so that the at least one processor 901 executes the above-mentioned big data-based abnormal number tracing method.

[0161] The specific implementation process of the processor 901 can refer to the above-mentioned method embodiments, which have similar implementation principles and technical effects, and will not be described here in detail.

[0162] In the above-mentioned Figure 9In the illustrated embodiment, it is to be understood that the processor can be a central processing unit (CPU), but can also be other general purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), etc. The general purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in connection with the application can be directly embodied as execution by a hardware processor, or be executed by a combination of hardware and software modules in the processor.

[0163] The memory can include a random access memory (RAM) and can also include a non-volatile memory (NVM), such as at least one disk memory.

[0164] The bus can be an industry standard architecture (ISA) bus, a peripheral component (PCI) bus, or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0165] The specific implementation of each operation can refer to the foregoing embodiments, which will not be described here.

[0166] According to an aspect of the present application, a computer program product or computer program is provided, which includes computer instructions stored in a computer readable storage medium.

[0167] It should be noted that for the foregoing method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited to the action sequence described, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0168] It should be understood that the above-mentioned apparatus embodiments are merely illustrative, and the apparatus of the present application can also be implemented in other manners. For example, the division of units / modules in the above-mentioned embodiments is merely a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units / modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0169] In addition, unless specifically stated otherwise, each functional unit / module in each embodiment of the present application can be integrated in one unit / module, or can be physically present separately, or two or more units / modules can be integrated together. The above-mentioned integrated unit / module can be realized in the form of hardware or in the form of a software program module.

[0170] In the above-mentioned embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments. Each technical feature of the above-mentioned embodiments can be combined arbitrarily, and in order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0171] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. The present application is intended to cover any variations, uses or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice in the art to which the application pertains or can relate. The specification and examples are to be regarded as exemplary only, and the true scope and spirit of the application are indicated by the following claims.

[0172] It should be understood that the present application is not limited to the precise construction that has been described above and illustrated in the accompanying drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the application is limited only by the claims that follow.

Claims

1. A method for processing credit card applications, characterized in that, include: Obtain the credit card information of a target user, wherein the target user is a user who has filled in credit card application information but has not submitted it, and the credit card information includes the target user's personal information, contact information and financial status; The target user's credit card information is analyzed using caching technology and data matching algorithms to determine the target user's intra-bank credit card information, which is used to indicate whether the user holds an intra-bank credit card or not. If the in-bank credit card information indicates that the target user does not hold an in-bank credit card, then target feature information is extracted from the credit card information according to the autoencoder algorithm. The target feature information includes the target user's consumption characteristics and needs. Based on the target feature information and matrix factorization algorithm, first service information is pushed to the target user, wherein the first service information is information on at least one target credit card determined according to the needs of the target user.

2. The method according to claim 1, characterized in that, The method further includes: If the intra-bank credit card information indicates that the target user holds an intra-bank credit card, then obtain the feature information of the intra-bank credit card held by the target user; Based on the feature information of the target user's credit card held by the bank and the logistic regression algorithm, the target user's satisfaction information with the credit card is determined; Based on the satisfaction information, second service information is pushed to the target user, the second service information including information on at least one adjustable service for the target user's credit card within the bank; The features of the credit cards mentioned above include credit card type, credit limit, usage, benefits, and annual fee. The satisfaction information includes satisfaction with credit card limits, satisfaction with benefits enjoyed, and satisfaction with annual fees.

3. The method according to claim 1 or 2, characterized in that, The target user's credit card information is analyzed using caching technology and data matching algorithms to determine the target user's intra-bank credit card information, including: The target user's credit card information is analyzed using caching technology to determine whether the target user is an identified user, where the identifier is the identifier of a user who already holds a credit card within the bank. If the target user is an identified user, then the target user's in-bank credit card information is determined to indicate that the user holds an in-bank credit card. If the target user is an unidentified user, the credit card information of the target user is analyzed according to the data matching algorithm to determine the target user's intra-bank credit card information.

4. The method according to claim 3, characterized in that, If the target user is an unidentified user, then the target user's credit card information is analyzed according to a data matching algorithm to determine the target user's intra-bank credit card information, including: The identity information of each user in the database is hashed to generate a hash value corresponding to each user's identity information. The database stores the identity information of each user holding a credit card within the bank, and each hash value corresponds one-to-one with each user. Based on the target user's credit card information, generate the target user's hash value to be compared; The hash value corresponding to each user's identity information is compared with the hash value to be compared for the target user to obtain the comparison result; If the comparison result indicates that there is a hash value in the database that matches the hash value to be compared of the target user, then it is determined that the target user holds an in-bank credit card. If the comparison result indicates that there is no hash value in the database that matches the hash value to be compared for the target user, then it is determined that the target user does not hold an internal credit card.

5. The method according to claim 1 or 2, characterized in that, If the intra-bank credit card information indicates that the target user does not hold an intra-bank credit card, then target feature information is extracted from the target user's credit card information according to the autoencoder algorithm, including: If the target user does not hold a credit card from the bank, the target user's credit card information is preprocessed to obtain the target user's feature information; Construct an autoencoder model based on the autoencoder algorithm; The autoencoder model is trained based on the feature information of the target user, and the target feature information is extracted from the feature information of the target user.

6. The method according to claim 1 or 2, characterized in that, Based on the target feature information and matrix factorization algorithm, first service information is pushed to the target user, including: Based on the target feature information and credit card item information, a user item interaction matrix is ​​constructed, wherein the credit card item information includes credit card product types; The target user item interaction matrix is ​​decomposed according to the matrix factorization algorithm to obtain the user feature matrix and the item feature matrix. By optimizing the learning algorithm and gradient descent method, the user feature matrix and the project feature matrix are optimized to determine the target user's demand for credit cards; Based on the target user's needs for credit cards, determine information on at least one target credit card; Based on the information of the at least one target credit card, push first service information to the target user.

7. The method according to claim 1 or 2, characterized in that, The method further includes: The target user's credit card information is stored on multiple nodes of the blockchain.

8. A credit card application processing device, characterized in that, include: The credit card information acquisition module is used to acquire the credit card information of a target user, wherein the target user is a user who has filled in credit card application information but has not submitted it, and the credit card information includes the target user's personal information, contact information and financial status. The in-bank credit card information determination module is used to analyze the credit card information of the target user using caching technology and data matching algorithm to determine the in-bank credit card information of the target user. The in-bank credit card information is used to indicate whether the user holds an in-bank credit card or not. The target feature information extraction module is used to extract target feature information from the credit card information according to the autoencoder algorithm if the in-bank credit card information indicates that the target user does not hold an in-bank credit card. The target feature information includes the target user's consumption characteristics and needs. The first service information push module is used to push first service information to the target user based on the target feature information and matrix factorization algorithm. The first service information is information about at least one target credit card determined according to the needs of the target user.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images