Automobile financial scheme matching method and system based on artificial intelligence

Abstract

The invention provides an automobile financial scheme matching method and system based on artificial intelligence, and relates to the technical field of artificial intelligence. In the application, first, target automobile financial data corresponding to a target automobile financial user and associated automobile financial data corresponding to an associated automobile financial user having an association relationship with the target automobile financial user are acquired; secondly, performing feature extraction operation on the target automobile financial data and the associated automobile financial data, and outputting target automobile financial feature representation and associated automobile financial feature representation; then, taking the associated automobile financial feature representation as an auxiliary feature representation of the target automobile financial feature representation, and analyzing financial scheme matching probability distribution; and finally, determining a target automobile financial scheme in the plurality of candidate automobile financial schemes based on the financial scheme matching probability distribution. Based on the above content, the problem that the reliability of automobile financial scheme matching is relatively low in the prior art can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to an artificial intelligence-based automobile finance solution matching method and system. Background Art

[0002] Automotive finance solutions refer to financial solutions related to the automotive business, such as auto loans, auto financing leases, auto insurance products, and auto value-preserving repurchase services. Among them, at points such as car purchases, corresponding auto finance solutions are generally recommended based on user needs. However, in the prior art, the corresponding staff generally conduct matching analysis based on the auto finance solutions they know, and then recommend the matching auto finance solutions to customers. However, the inventors have found through research that when the staff conducts matching of auto solutions, the reliability of the matching analysis is relatively low. Summary of the invention

[0003] In view of this, the purpose of the present application is to provide an automobile finance plan matching method and system based on artificial intelligence to improve the problem of relatively low reliability of automobile finance plan matching in the prior art.

[0004] To achieve the above objectives, this application adopts the following technical solutions:

[0005] An automobile financial solution matching method based on artificial intelligence, comprising:

[0006] Acquire target automobile finance data corresponding to a target automobile finance user, and acquire associated automobile finance data corresponding to an associated automobile finance user having an associated relationship with the target automobile finance user, wherein the target automobile finance data at least includes user identity information and user property information of the target automobile finance user, and the associated automobile finance data at least includes user identity information and user property information of the associated automobile finance user;

[0007] Performing feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and outputting a target automobile finance feature representation corresponding to the target automobile finance data and an associated automobile finance feature representation corresponding to the associated automobile finance data;

[0008] The associated automobile finance feature representation is used as an auxiliary feature representation of the target automobile finance feature representation, and a finance plan matching probability distribution corresponding to the target automobile finance user is analyzed and output, wherein the finance plan matching probability distribution includes a plurality of finance plan matching probabilities respectively used to characterize the matching probabilities of a plurality of candidate automobile finance plans;

[0009] Based on the finance plan matching probability distribution, a target automobile finance plan corresponding to the target automobile finance user is determined from among the multiple candidate automobile finance plans.

[0010] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, the step of obtaining target automobile finance data corresponding to the target automobile finance user, and obtaining associated automobile finance data corresponding to associated automobile finance users having an associated relationship with the target automobile finance user, comprises:

[0011] Obtain target automobile finance data corresponding to the target automobile finance user;

[0012] In the pre-built automobile finance user set, at least one other automobile finance user having interaction with the target automobile finance user is determined, and the other automobile finance user is determined as an associated automobile finance user having an associated relationship with the target automobile finance user;

[0013] The associated automobile finance data corresponding to the associated automobile finance user is obtained.

[0014] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, the step of performing feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and outputting the target automobile finance feature representation corresponding to the target automobile finance data and the associated automobile finance feature representation corresponding to the associated automobile finance data, comprises:

[0015] Performing feature extraction operations on the user identity information and the user property information in the target automobile financial data, respectively, and outputting a target user identity feature representation and a target user property feature representation corresponding to the target automobile financial data;

[0016] Performing feature extraction operations on the user identity information, user property information and selected automobile finance scheme in the associated automobile finance data, and outputting associated user identity feature representation, associated user property feature representation and associated finance scheme feature representation corresponding to the associated automobile finance data;

[0017] fusing the target user identity feature representation and the target user property feature representation to form a target automobile financial feature representation corresponding to the target automobile financial data;

[0018] fusing the associated user identity feature representation and the associated user property feature representation to form a fused automobile finance feature representation corresponding to the target automobile finance data;

[0019] Based on the associated financial solution feature representation, a semantic enhancement operation is performed on the fused automobile finance feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

[0020] In a preferred embodiment of the present application, in the above-mentioned automobile finance scheme matching method based on artificial intelligence, the step of performing a semantic enhancement operation on the fused automobile finance feature representation based on the associated finance scheme feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data includes:

[0021] In the semantic enhancement operation at the first depth, semantic information having correlation with the associated financial scheme feature representation is mined from the fused automobile finance feature representation to obtain a corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain an associated automobile finance feature representation at the first depth;

[0022] In the semantic enhancement operation at other depths other than the first depth, the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation at the previous depth are downsampled to achieve compression of the feature representation, and the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation at the current depth are obtained, and semantic information having correlation with the associated financial scheme feature representation in the semantic enhancement operation at the current depth is mined from the fused automobile finance feature representation in the semantic enhancement operation at the current depth to obtain the corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation in the semantic enhancement operation at the current depth to achieve semantic enhancement of the fused automobile finance feature representation in the semantic enhancement operation at the current depth to obtain the associated automobile finance feature representation at the current depth;

[0023] After the last depth of the associated automobile finance feature representation is obtained, the associated automobile finance feature representations at each depth are fused to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

[0024] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, in the semantic enhancement operation at the first depth, the semantic information having correlation with the associated financial solution feature representation is mined from the fused automobile finance feature representation to obtain the corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain the associated automobile finance feature representation at the first depth, including:

[0025] In a semantic enhancement operation of the first depth, a correlation parameter distribution between the fused automobile finance feature representation and the associated financial scheme feature representation is determined, wherein the correlation parameter distribution is obtained by multiplying the associated financial scheme feature representation and the fused automobile finance feature representation;

[0026] Based on the correlation parameter distribution, weighting the fused automobile finance feature representation to obtain a corresponding correlation feature representation, wherein a parameter in the correlation feature representation is equal to a product of the correlation parameter distribution and a parameter at a corresponding position in the fused automobile finance feature representation;

[0027] The correlation feature representation and the fused automobile finance feature representation are added or averaged to achieve semantic enhancement of the fused automobile finance feature representation, thereby obtaining a correlation automobile finance feature representation of a first depth, wherein a parameter in the correlation automobile finance feature representation is equal to a sum or average of parameters at corresponding positions in the correlation feature representation and the fused automobile finance feature representation.

[0028] In a preferred embodiment of the present application, in the above-mentioned automobile finance scheme matching method based on artificial intelligence, the step of using the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, and analyzing and outputting the financial scheme matching probability distribution corresponding to the target automobile finance user, comprises:

[0029] Extracting semantic information related to the associated automobile finance feature representation from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation;

[0030] A full connection operation is performed on the target financial solution feature representation to output the corresponding fully connected financial solution feature, and based on the fully connected financial solution feature, a financial solution matching probability distribution corresponding to the target automobile financial user is analyzed and output.

[0031] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, the steps of mining semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation include:

[0032] Performing multi-depth correlation mining on the target automobile finance feature representation and the associated automobile finance feature representation, so as to mine semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation, and obtain corresponding correlation feature representation;

[0033] The correlation feature representation is fused with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation and obtain a corresponding target finance solution feature representation.

[0034] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, the step of performing multi-depth correlation mining on the target automobile finance feature representation and the associated automobile finance feature representation to mine semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation to obtain the corresponding correlation feature representation includes:

[0035] In the first depth correlation mining, semantic information having correlation with the associated automobile finance feature representation is mined from the target automobile finance feature representation to obtain the first depth correlation feature representation;

[0036] In the correlation mining at other depths other than the first depth, a local masking operation is performed on the correlation feature representation of the previous depth to obtain a local masked feature representation of the current depth, and semantic information having correlation with the local masked feature representation of the current depth is mined from the target automobile finance feature representation to obtain a correlation feature representation of the current depth;

[0037] After obtaining the correlation feature representation of the last depth, the correlation feature representations of each depth are fused to form the final correlation feature representation.

[0038] In a preferred embodiment of the present application, in the above-mentioned artificial intelligence-based automobile finance solution matching method, the artificial intelligence-based automobile finance solution matching method further includes:

[0039] Acquire first automobile finance data corresponding to a first automobile finance user, and acquire second automobile finance data corresponding to a second automobile finance user associated with the first automobile finance user, wherein the first automobile finance data at least includes user identity information and user property information of the first automobile finance user, and the second automobile finance data at least includes user identity information and user property information of the second automobile finance user, and the first automobile finance data and the second automobile finance data are used as sample data for training a neural network;

[0040] Using a feature extraction model included in a candidate financial solution matching network, feature extraction operations are performed on the first automobile financial data and the second automobile financial data respectively, and a first automobile financial feature representation corresponding to the first automobile financial data and a second automobile financial feature representation corresponding to the second automobile financial data are output, wherein the candidate financial solution matching network is a neural network;

[0041] Using the financial solution matching model included in the candidate financial solution matching network, taking the second automobile financial feature representation as an auxiliary feature representation of the first automobile financial feature representation, and analyzing and outputting a financial solution matching probability distribution corresponding to the first automobile financial user;

[0042] Based on the financial solution matching probability distribution corresponding to the first automobile finance user and the actually selected automobile finance solution corresponding to the first automobile finance user, a training error of the candidate financial solution matching network is analyzed, and based on the training error, a network parameter included in the candidate financial solution matching network is updated to form a target financial solution matching network corresponding to the candidate financial solution matching network, wherein the target financial solution matching network is used to analyze and output the target automobile finance solution corresponding to the target automobile finance user.

[0043] On the basis of the above, the present application also provides an artificial intelligence-based automobile financial solution matching system, including:

[0044] Memory for storing computer programs;

[0045] The processor connected to the memory is used to execute the computer program stored in the memory to implement the above-mentioned automobile financial solution matching method based on artificial intelligence.

[0046] The artificial intelligence-based automobile finance scheme matching method and system provided in the present application firstly obtain target automobile finance data corresponding to a target automobile finance user and associated automobile finance data corresponding to associated automobile finance users having an associated relationship with the target automobile finance user; secondly, perform feature extraction operations on the target automobile finance data and the associated automobile finance data, and output a target automobile finance feature representation and an associated automobile finance feature representation; then, use the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, and analyze a finance scheme matching probability distribution; finally, determine a target automobile finance scheme from a plurality of candidate automobile finance schemes based on the finance scheme matching probability distribution. Based on the above content, since when conducting financial scheme matching analysis, the basis of the analysis includes not only the target automobile finance feature representation corresponding to the target automobile finance data, but also the associated automobile finance feature representation corresponding to the associated automobile finance data, the basis of the analysis can be more sufficient and comprehensive. In addition, since there is an association relationship between the associated automobile finance users and the target automobile finance users, that is, the associated automobile finance data has a certain reinforcing effect on the target automobile finance data, therefore, the matching analysis based on the two aspects of data is reliable, so that the reliability of the determined target automobile finance scheme can be effectively improved, thereby improving the problem of relatively low reliability of automobile finance scheme matching existing in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings.

[0048] Figure 1 A structural block diagram of an artificial intelligence-based automobile financial solution matching system provided in an embodiment of the present application.

[0049] Figure 2 A flowchart of an artificial intelligence-based automobile finance solution matching method provided in an embodiment of the present application.

[0050] Figure 3 A schematic diagram of a multi-depth semantic enhancement operation provided in an embodiment of the present application.

[0051] Figure 4 A schematic diagram of multi-depth correlation mining provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0053] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0054] like Figure 1 As shown, an embodiment of the present application provides an automobile financial solution matching system based on artificial intelligence. Among them, the automobile financial solution matching system based on artificial intelligence may include a memory and a processor. In detail, the memory and the processor are directly or indirectly electrically connected to realize data transmission or interaction. For example, the memory and the processor can be electrically connected through one or more communication buses or signal lines. The processor is used to execute an executable computer program stored in the memory to implement the automobile financial solution matching method based on artificial intelligence provided in the embodiment of the present application.

[0055] Optionally, the memory can be, but is not limited to, random access memory (RAM), read only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable read-only memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc.

[0056] Furthermore, the processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0057] Understandably, Figure 1 The structure shown is for illustration only. The automobile finance solution matching system based on artificial intelligence may also include Figure 1 More or fewer components as shown, or with Figure 1 The different configurations shown, for example, may also include a communication unit for information exchange with other devices. In addition, the automobile financial solution matching system based on artificial intelligence may be a server.

[0058] Combination Figure 2 The embodiment of the present application also provides an artificial intelligence-based automobile financial solution matching method that can be applied to the above-mentioned artificial intelligence-based automobile financial solution matching system. Among them, the method steps defined in the process related to the artificial intelligence-based automobile financial solution matching method can be implemented by the artificial intelligence-based automobile financial solution matching system (hereinafter referred to as the solution matching system). Figure 2 The specific process shown is explained in detail.

[0059] Step S110 , obtaining target automobile finance data corresponding to a target automobile finance user, and obtaining associated automobile finance data corresponding to associated automobile finance users having an associated relationship with the target automobile finance user.

[0060] In the embodiment of the present application, the solution matching system can obtain the target automobile finance data corresponding to the target automobile finance user, and obtain the associated automobile finance data corresponding to the associated automobile finance users who have an associated relationship with the target automobile finance user. The target automobile finance data at least includes the user identity information (such as gender, age, occupation, family situation, etc.) and user property information (such as income, assets, credit score, etc.) of the target automobile finance user, and the associated automobile finance data at least includes the user identity information and user property information of the associated automobile finance user; illustratively, in other implementations, other information may also be included.

[0061] Step S120 , performing feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and outputting target automobile finance feature representations corresponding to the target automobile finance data and associated automobile finance feature representations corresponding to the associated automobile finance data.

[0062] In the embodiment of the present application, after obtaining the target automobile finance data and the associated automobile finance data, the solution matching system can perform feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and output the target automobile finance feature representation corresponding to the target automobile finance data and the associated automobile finance feature representation corresponding to the associated automobile finance data. That is, the target automobile finance feature representation can be used to reflect the semantic information in the target automobile finance data, and the associated automobile finance feature representation can be used to reflect the semantic information in the associated automobile finance data. In addition, in the embodiment of the present application, the specific expression form of each feature representation can be a vector or a matrix.

[0063] Step S130 , using the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, analyzing and outputting a financial plan matching probability distribution corresponding to the target automobile finance user.

[0064] In an embodiment of the present application, after obtaining the target automobile finance feature representation and the associated automobile finance feature representation, the solution matching system may use the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, and analyze and output the financial solution matching probability distribution corresponding to the target automobile finance user. Among them, the multiple financial solution matching probabilities included in the financial solution matching probability distribution are respectively used to characterize the matching probabilities of multiple candidate automobile finance solutions, such as the matching probability of candidate automobile finance solution 1, the matching probability of candidate automobile finance solution 2, and the matching probability of candidate automobile finance solution 3. In addition, the candidate automobile finance solution may refer to an automobile insurance solution, which may include vehicle loss insurance, commercial third-party liability insurance, vehicle occupant liability insurance, non-deductible special insurance, scratch insurance, spontaneous combustion loss insurance, etc., and each insurance type also has corresponding information such as the amount.

[0065] Step S140 : determining a target automobile finance plan corresponding to the target automobile finance user from among the multiple candidate automobile finance plans based on the finance plan matching probability distribution.

[0066] In an embodiment of the present application, after obtaining the financial plan matching probability distribution, the plan matching system can determine the target automobile finance plan corresponding to the target automobile finance user from the multiple candidate automobile finance plans based on the financial plan matching probability distribution. Exemplarily, the candidate automobile finance plan corresponding to the financial plan matching probability with the maximum value in the financial plan matching probability distribution can be determined as the target automobile finance plan.

[0067] Based on the above content, since when conducting financial scheme matching analysis, the basis of the analysis includes not only the target automobile finance feature representation corresponding to the target automobile finance data, but also the associated automobile finance feature representation corresponding to the associated automobile finance data, the basis of the analysis can be more sufficient and comprehensive. In addition, since there is an association relationship between the associated automobile finance users and the target automobile finance users, that is, the associated automobile finance data has a certain reinforcing effect on the target automobile finance data, therefore, the matching analysis based on the two aspects of data is reliable, so that the reliability of the determined target automobile finance scheme can be effectively improved, thereby improving the problem of relatively low reliability of automobile finance scheme matching existing in the prior art.

[0068] It should be noted that for step S110 , the specific method of obtaining the corresponding automobile financial data is not limited and can be selected according to actual needs.

[0069] For example, in an alternative implementation, in order to ensure that the obtained automobile financial data has a high reliability, the above step S110 may further include the following content:

[0070] First, target automobile finance data corresponding to the target automobile finance user may be obtained, wherein the target automobile finance data may be formed by filling in the data by the target automobile finance user;

[0071] Secondly, at least one other automobile finance user (which may be the closest other automobile finance user) who has an interaction with the target automobile finance user (such as an interaction relationship determined based on the collected information, such as a family or friend relationship) may be determined from a pre-built automobile finance user set (which may be a set of users who agree to collect and apply corresponding information), and the other automobile finance user may be determined as an associated automobile finance user who has an associated relationship with the target automobile finance user;

[0072] Finally, the associated automobile finance data corresponding to the associated automobile finance user may be obtained.

[0073] It should be noted that for step S120 , the specific manner of performing feature extraction operations on the target automobile financial data and the associated automobile financial data is not limited and can be selected according to needs.

[0074] For example, in an alternative implementation, in order to ensure that the extracted target automobile financial feature representation and the associated automobile financial feature representation have better semantic representation capabilities, the above-mentioned step S120 may further include step S121, step S122, step S123, step S124 and step S125, and the specific content of each step is described as follows.

[0075] Step S121 , performing feature extraction operations on the user identity information and the user property information in the target automobile financial data, respectively, and outputting target user identity feature representation and target user property feature representation corresponding to the target automobile financial data.

[0076] In an embodiment of the present application, feature extraction operations can be performed on the user identity information and user property information in the target automobile financial data, respectively, and the target user identity feature representation and target user property feature representation corresponding to the target automobile financial data can be output. In other words, feature extraction operations can be performed on the user identity information to output the target user identity feature representation, and feature extraction operations can be performed on the user property information to output the target user property feature representation. In addition, in a specific implementation, word embedding processing can be performed based on a word embedding model to complete the feature extraction operation. For example, the user identity information "gender is male" is segmented and embedded, and the following can be obtained:

[0077] "gender": [0.23, -0.1, 0.17, 0.32, -0.08, 0.45, -0.12, 0.29, -0.33, 0.11, 0.24, -0.05, ..., 0.17];

[0078] "for": [0.12, -0.33, 0.22, 0.44, 0.01, 0.23, 0.14, -0.19, 0.37, -0.24, 0.16, 0.22, ..., 0.09];

[0079] “Male” : [0.04, 0.18, -0.11, -0.28, 0.33, -0.12, 0.05, -0.21, 0.19, 0.07, -0.02, 0.14, …, 0.32].

[0080] Based on this, the embedding vectors corresponding to each word in the user identity information can be averaged to obtain the target user identity feature representation.

[0081] Step S122, performing feature extraction operations on the user identity information, user property information and selected automobile financial scheme in the associated automobile financial data, and outputting the associated user identity feature representation, associated user property feature representation and associated financial scheme feature representation corresponding to the associated automobile financial data.

[0082] In an embodiment of the present application, feature extraction operations (as described above) can be performed on the user identity information, user property information and selected automobile finance plan (i.e., automobile finance plan adopted historically) in the associated automobile finance data, and the associated user identity feature representation, associated user property feature representation and associated finance plan feature representation corresponding to the associated automobile finance data are output.

[0083] Step S123: fusing the target user identity feature representation and the target user property feature representation to form a target automobile financial feature representation corresponding to the target automobile financial data.

[0084] In the embodiment of the present application, after obtaining the target user identity feature representation and the target user property feature representation, the target user identity feature representation and the target user property feature representation can be fused to form a target automobile financial feature representation corresponding to the target automobile financial data, so that the target automobile financial feature representation has two aspects of semantic information. Exemplarily, the target user identity feature representation and the target user property feature representation can be added or averaged to obtain the target automobile financial feature representation.

[0085] Step S124, fusing the associated user identity feature representation and the associated user property feature representation to form a fused automobile finance feature representation corresponding to the target automobile finance data.

[0086] In the embodiment of the present application, after obtaining the associated user identity feature representation, the associated user property feature representation and the associated financial scheme feature representation, the associated user identity feature representation and the associated user property feature representation may be fused to form a fused automobile financial feature representation corresponding to the target automobile financial data, so that the fused automobile financial feature representation has two aspects of semantic information. Exemplarily, the associated user identity feature representation and the associated user property feature representation may be added or averaged to obtain the fused automobile financial feature representation.

[0087] Step S125 , based on the associated financial solution feature representation, a semantic enhancement operation is performed on the fused automobile finance feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

[0088] In an embodiment of the present application, after forming a fused automobile finance feature representation, a semantic enhancement operation can be further performed on the fused automobile finance feature representation based on the associated financial scheme feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data. In this way, the associated automobile finance feature representation can have three aspects of semantic information.

[0089] It is understandable that in the above step S124, the specific manner of performing semantic enhancement operation on the fused auto finance feature representation is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to achieve reliable semantic enhancement operation so that the obtained associated auto finance feature representation can effectively characterize the information of the associated auto finance user, the above step S124 can further include step S124a, step S124b and step S124c, and the specific contents of each step are as follows (combined with Figure 3 shown).

[0090] Step S124a, in the semantic enhancement operation at the first depth, semantic information having correlation with the associated financial scheme feature representation is mined from the fused automobile finance feature representation to obtain a corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain an associated automobile finance feature representation at the first depth.

[0091] In the embodiment of the present application, in the semantic enhancement operation at the first depth, semantic information having correlation with the associated financial scheme feature representation can be mined from the fused automobile finance feature representation to obtain the corresponding correlation feature representation, and the correlation feature representation can be fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain the associated automobile finance feature representation at the first depth. Based on this, the associated automobile finance feature representation at the first depth can represent both the semantic information of the mined correlation and the semantic information of the fused automobile finance feature representation itself, thereby avoiding the problem of losing its own important semantic information when mining the semantic information of the correlation.

[0092] Step S124b, in the semantic enhancement operation at other depths other than the first depth, downsample the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation of the previous depth to achieve compression of the feature representation, and obtain the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation of the current depth, and mine the semantic information that is correlated with the associated financial scheme feature representation in the semantic enhancement operation of the current depth from the fused automobile finance feature representation in the semantic enhancement operation of the current depth to obtain the corresponding correlation feature representation, and fuse the correlation feature representation with the fused automobile finance feature representation in the semantic enhancement operation of the current depth to achieve semantic enhancement of the fused automobile finance feature representation in the semantic enhancement operation of the current depth to obtain the associated automobile finance feature representation of the current depth.

[0093] In an embodiment of the present application, in semantic enhancement operations at depths other than the first depth (such as semantic enhancement operations at the second depth and semantic enhancement operations at the third depth, etc.), the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation of the previous depth may be downsampled (the specific sampling parameters of the downsampling may be configured according to actual needs) to achieve compression of the feature representation, and obtain the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation of the current depth, and from the fused automobile finance feature representation in the semantic enhancement operation of the current depth, semantic information having correlation with the associated financial scheme feature representation in the semantic enhancement operation of the current depth may be mined to obtain the corresponding correlation feature representation, and the correlation feature representation may be fused with the fused automobile finance feature representation in the semantic enhancement operation of the current depth to achieve semantic enhancement of the fused automobile finance feature representation in the semantic enhancement operation of the current depth, and obtain the associated automobile finance feature representation of the current depth.

[0094] Step S124c: after obtaining the last depth of the associated automobile finance feature representation, the associated automobile finance feature representations at each depth are merged to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

[0095] In the embodiment of the present application, after obtaining the associated automobile finance feature representation of the last depth, the associated automobile finance feature representations of each depth may be fused (such as added or averaged) to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

[0096] It is understandable that, in the above-mentioned steps S124a and S124b, the specific method of mining the semantic information of the correlation is not limited (the specific methods of the two steps can be the same), and can be selected according to actual needs. For example, in an alternative implementation, the above-mentioned step S124a can further include the following specific implementation content:

[0097] First, in a semantic enhancement operation at a first depth, a correlation parameter distribution between the fused automobile finance feature representation and the associated financial solution feature representation may be determined, wherein the correlation parameter distribution is obtained by multiplying the associated financial solution feature representation and (the transposed result of) the fused automobile finance feature representation;

[0098] Secondly, the fused automobile finance feature representation may be weighted based on the correlation parameter distribution to obtain a corresponding correlation feature representation, wherein the parameter in the correlation feature representation is equal to the product of the correlation parameter distribution and the parameter at the corresponding position in the fused automobile finance feature representation, that is, bitwise multiplication is performed;

[0099] Then, the correlation feature representation and the fused automobile finance feature representation may be added or averaged to achieve semantic enhancement of the fused automobile finance feature representation, and obtain a first depth of associated automobile finance feature representation, wherein the parameters in the associated automobile finance feature representation are equal to the sum or average of the parameters at corresponding positions in the correlation feature representation and the fused automobile finance feature representation, that is, bitwise addition or average calculation is performed.

[0100] It should be noted that for step S130 , the specific method of analyzing and outputting the financial plan matching probability distribution corresponding to the target automobile financial user is not limited and can be selected according to actual needs.

[0101] For example, in an alternative implementation, in order to ensure the reliability of the analysis output of the financial solution matching probability distribution, the above-mentioned step S130 may further include step S131 and step S132, and the specific implementation content of each step is described as follows.

[0102] Step S131, mining semantic information that is correlated with the associated automobile finance feature representation from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation.

[0103] In the embodiment of the present application, semantic information having correlation with the associated automobile finance feature representation can be mined from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and the correlation feature representation can be fused with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation. In other words, the target automobile finance feature representation can be semantically enhanced based on the associated automobile finance feature representation.

[0104] Step S132, performing a full connection operation on the target financial solution feature representation, outputting the corresponding fully connected financial solution feature, and, based on the fully connected financial solution feature, analyzing and outputting the financial solution matching probability distribution corresponding to the target automobile financial user.

[0105] In an embodiment of the present application, after obtaining the target financial solution feature representation, a full connection operation can be performed on the target financial solution feature representation to output the corresponding fully connected financial solution feature, and based on the fully connected financial solution feature, the financial solution matching probability distribution corresponding to the target automobile finance user is analyzed and output. Exemplarily, the fully connected financial solution feature can be processed by a classification function such as softmax to obtain the matching probability of each candidate automobile finance solution, that is, the financial solution matching probability distribution is obtained.

[0106] It is understandable that, in the above step S131, the specific method of mining and fusing the semantic information of correlation is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to ensure the reliability of mining and fusing the semantic information of correlation so that the target financial solution feature representation can fully and reliably characterize the matching candidate automobile financial solution, the above step S131 can further include step S131a and step S131b, and the specific content of each step is described as follows.

[0107] Step S131a, performing multi-depth correlation mining on the target automobile finance feature representation and the associated automobile finance feature representation, so as to mine semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation, and obtain corresponding correlation feature representation.

[0108] In an embodiment of the present application, the target automobile finance feature representation and the associated automobile finance feature representation may be subjected to multi-depth correlation mining (to capture more detailed information) to mine semantic information that is correlated with the associated automobile finance feature representation from the target automobile finance feature representation, and obtain a corresponding correlation feature representation.

[0109] Step S131b, fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation and obtain a corresponding target finance solution feature representation.

[0110] In the embodiment of the present application, after obtaining the correlation feature representation, the correlation feature representation and the target automobile finance feature representation can be fused to achieve semantic enhancement of the target automobile finance feature representation and obtain the corresponding target finance solution feature representation. Exemplarily, the correlation feature representation and the target automobile finance feature representation can be added or averaged to obtain a fully fused target finance solution feature representation.

[0111] It can be understood that in the above step S131a, the specific method of performing multi-depth correlation mining is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to ensure the reliability of multi-depth correlation mining, so that the obtained correlation feature representation can have a more reliable semantic representation ability, the above step S131a can further include the following specific implementation content (combined with Figure 4 shown):

[0112] First, in the first depth correlation mining, semantic information having correlation with the associated automobile finance feature representation can be mined from the target automobile finance feature representation to obtain the first depth correlation feature representation. The specific mining process can refer to the related description above.

[0113] Secondly, in the correlation mining of other depths other than the first depth (such as the correlation mining of the second depth and the correlation mining of the third depth, etc.), the correlation feature representation of the previous depth can be subjected to a local masking operation (for example, it can be bitwise multiplied with a masking matrix, the masking matrix can be used as a parameter of the corresponding neural network and is formed during the training process, and the parameters of the masking matrix are composed of parameters 1 and 0; in addition, in order to avoid introducing excessive interference, the number of parameters 0 can be much smaller than the number of parameters 1; in addition, the masking matrices corresponding to the local masking operations in the correlation mining of different depths can be different), to obtain the local masked feature representation of the current depth, and to mine the semantic information that is correlated with the local masked feature representation of the current depth from the target automobile finance feature representation, to obtain the correlation feature representation of the current depth;

[0114] Finally, after obtaining the correlation feature representation of the last depth, the correlation feature representations of each depth may be fused to form a final correlation feature representation; illustratively, the correlation feature representations of each depth may be added or averaged.

[0115] It should also be noted that for steps S110 to S140, in order to reliably determine the matching target automobile financial solution, the automobile financial solution matching method based on artificial intelligence may further include the step of training a target financial solution matching network, which may specifically include:

[0116] First, first automobile finance data corresponding to a first automobile finance user may be obtained, and second automobile finance data corresponding to a second automobile finance user associated with the first automobile finance user may be obtained, wherein the first automobile finance data at least includes user identity information and user property information of the first automobile finance user, and the second automobile finance data at least includes user identity information and user property information of the second automobile finance user, and the first automobile finance data and the second automobile finance data are used as sample data for training a neural network;

[0117] Secondly, the feature extraction model included in the candidate financial solution matching network can be used to perform feature extraction operations on the first automobile financial data and the second automobile financial data, respectively, and output a first automobile financial feature representation corresponding to the first automobile financial data and a second automobile financial feature representation corresponding to the second automobile financial data, wherein the candidate financial solution matching network is a neural network; in addition, the specific processing method of the feature extraction operation can refer to the above description;

[0118] Then, the financial solution matching model included in the candidate financial solution matching network can be used to analyze and output the financial solution matching probability distribution corresponding to the first automobile financial user by using the second automobile financial feature representation as an auxiliary feature representation of the first automobile financial feature representation, wherein the specific process of the analysis and output can refer to the relevant description above;

[0119] Finally, based on the financial solution matching probability distribution corresponding to the first automobile finance user and the actually selected automobile finance solution corresponding to the first automobile finance user, the training error of the candidate financial solution matching network can be analyzed (exemplarily, mean square error calculation can be performed to obtain the corresponding training error), and based on the training error, the network parameters included in the candidate financial solution matching network are updated to form a target financial solution matching network corresponding to the candidate financial solution matching network (specifically, the network parameters can be updated in a direction of reducing the training error until the training error converges), wherein the target financial solution matching network is used to analyze and output the target automobile finance solution corresponding to the target automobile finance user.

[0120] In summary, the artificial intelligence-based automobile finance scheme matching method and system provided by the present application firstly obtains target automobile finance data corresponding to a target automobile finance user and associated automobile finance data corresponding to associated automobile finance users having an associated relationship with the target automobile finance user; secondly, performs feature extraction operations on the target automobile finance data and the associated automobile finance data, and outputs a target automobile finance feature representation and an associated automobile finance feature representation; then, uses the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, and analyzes a financial scheme matching probability distribution; finally, determines a target automobile finance scheme from a plurality of candidate automobile finance schemes based on the financial scheme matching probability distribution. Based on the above content, since when conducting financial scheme matching analysis, the basis of the analysis includes not only the target automobile finance feature representation corresponding to the target automobile finance data, but also the associated automobile finance feature representation corresponding to the associated automobile finance data, the basis of the analysis can be more sufficient and comprehensive. In addition, since there is an association relationship between the associated automobile finance users and the target automobile finance users, that is, the associated automobile finance data has a certain reinforcing effect on the target automobile finance data, therefore, the matching analysis based on the two aspects of data is reliable, so that the reliability of the determined target automobile finance scheme can be effectively improved, thereby improving the problem of relatively low reliability of automobile finance scheme matching existing in the prior art.

[0121] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device and method embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the device, method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0122] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0123] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk. It should be noted that, in this article, the term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0124] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An automobile finance solution matching method based on artificial intelligence, characterized in that: include: Acquire target automobile finance data corresponding to a target automobile finance user, and acquire associated automobile finance data corresponding to an associated automobile finance user having an associated relationship with the target automobile finance user, wherein the target automobile finance data at least includes user identity information and user property information of the target automobile finance user, and the associated automobile finance data at least includes user identity information and user property information of the associated automobile finance user; Performing feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and outputting a target automobile finance feature representation corresponding to the target automobile finance data and an associated automobile finance feature representation corresponding to the associated automobile finance data; The associated automobile finance feature representation is used as an auxiliary feature representation of the target automobile finance feature representation, and a finance plan matching probability distribution corresponding to the target automobile finance user is analyzed and output, wherein the finance plan matching probability distribution includes a plurality of finance plan matching probabilities respectively used to characterize the matching probabilities of a plurality of candidate automobile finance plans; Based on the finance plan matching probability distribution, a target automobile finance plan corresponding to the target automobile finance user is determined from among the multiple candidate automobile finance plans.

2. The artificial intelligence-based automobile finance solution matching method according to claim 1, characterized in that: The step of acquiring target automobile finance data corresponding to the target automobile finance user, and acquiring associated automobile finance data corresponding to associated automobile finance users having an associated relationship with the target automobile finance user, comprises: Obtain target automobile finance data corresponding to the target automobile finance user; In the pre-built automobile finance user set, at least one other automobile finance user having interaction with the target automobile finance user is determined, and the other automobile finance user is determined as an associated automobile finance user having an associated relationship with the target automobile finance user; The associated automobile finance data corresponding to the associated automobile finance user is obtained.

3. The automobile finance solution matching method based on artificial intelligence according to claim 1 is characterized in that: The step of performing feature extraction operations on the target automobile finance data and the associated automobile finance data respectively, and outputting a target automobile finance feature representation corresponding to the target automobile finance data and an associated automobile finance feature representation corresponding to the associated automobile finance data, comprises: Performing feature extraction operations on the user identity information and the user property information in the target automobile financial data, respectively, and outputting a target user identity feature representation and a target user property feature representation corresponding to the target automobile financial data; Performing feature extraction operations on the user identity information, user property information and selected automobile finance scheme in the associated automobile finance data, and outputting associated user identity feature representation, associated user property feature representation and associated finance scheme feature representation corresponding to the associated automobile finance data; fusing the target user identity feature representation and the target user property feature representation to form a target automobile financial feature representation corresponding to the target automobile financial data; fusing the associated user identity feature representation and the associated user property feature representation to form a fused automobile finance feature representation corresponding to the target automobile finance data; Based on the associated financial solution feature representation, a semantic enhancement operation is performed on the fused automobile finance feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

4. The artificial intelligence-based automobile finance solution matching method according to claim 3 is characterized in that: The step of performing a semantic enhancement operation on the fused automobile finance feature representation based on the associated finance scheme feature representation to form an associated automobile finance feature representation corresponding to the associated automobile finance data includes: In the semantic enhancement operation at the first depth, semantic information having correlation with the associated financial scheme feature representation is mined from the fused automobile finance feature representation to obtain a corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain an associated automobile finance feature representation at the first depth; In the semantic enhancement operation at other depths other than the first depth, the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation at the previous depth are downsampled to achieve compression of the feature representation, and the fused automobile finance feature representation and the associated financial scheme feature representation in the semantic enhancement operation at the current depth are obtained, and semantic information having correlation with the associated financial scheme feature representation in the semantic enhancement operation at the current depth is mined from the fused automobile finance feature representation in the semantic enhancement operation at the current depth to obtain the corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation in the semantic enhancement operation at the current depth to achieve semantic enhancement of the fused automobile finance feature representation in the semantic enhancement operation at the current depth to obtain the associated automobile finance feature representation at the current depth; After the last depth of the associated automobile finance feature representation is obtained, the associated automobile finance feature representations at each depth are fused to form an associated automobile finance feature representation corresponding to the associated automobile finance data.

5. The artificial intelligence-based automobile finance solution matching method according to claim 4 is characterized in that: In the semantic enhancement operation at the first depth, the semantic information having correlation with the associated financial scheme feature representation is mined from the fused automobile finance feature representation to obtain a corresponding correlation feature representation, and the correlation feature representation is fused with the fused automobile finance feature representation to achieve semantic enhancement of the fused automobile finance feature representation to obtain the associated automobile finance feature representation at the first depth, including: In a semantic enhancement operation of the first depth, a correlation parameter distribution between the fused automobile finance feature representation and the associated financial scheme feature representation is determined, wherein the correlation parameter distribution is obtained by multiplying the associated financial scheme feature representation and the fused automobile finance feature representation; Based on the correlation parameter distribution, weighting the fused automobile finance feature representation to obtain a corresponding correlation feature representation, wherein a parameter in the correlation feature representation is equal to a product of the correlation parameter distribution and a parameter at a corresponding position in the fused automobile finance feature representation; The correlation feature representation and the fused automobile finance feature representation are added or averaged to achieve semantic enhancement of the fused automobile finance feature representation, thereby obtaining a correlation automobile finance feature representation of a first depth, wherein a parameter in the correlation automobile finance feature representation is equal to a sum or average of parameters at corresponding positions in the correlation feature representation and the fused automobile finance feature representation.

6. The artificial intelligence-based automobile finance solution matching method according to claim 1, characterized in that: The step of using the associated automobile finance feature representation as an auxiliary feature representation of the target automobile finance feature representation, and analyzing and outputting a matching probability distribution of a financial plan corresponding to the target automobile finance user, comprises: Extracting semantic information related to the associated automobile finance feature representation from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation; A full connection operation is performed on the target financial solution feature representation to output the corresponding fully connected financial solution feature, and based on the fully connected financial solution feature, a financial solution matching probability distribution corresponding to the target automobile financial user is analyzed and output.

7. The artificial intelligence-based automobile finance solution matching method according to claim 6, characterized in that: The step of mining semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation to obtain a corresponding correlation feature representation, and fusing the correlation feature representation with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation to obtain a corresponding target finance solution feature representation includes: Performing multi-depth correlation mining on the target automobile finance feature representation and the associated automobile finance feature representation, so as to mine semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation, and obtain corresponding correlation feature representation; The correlation feature representation is fused with the target automobile finance feature representation to achieve semantic enhancement of the target automobile finance feature representation and obtain a corresponding target finance solution feature representation.

8. The artificial intelligence-based automobile finance solution matching method according to claim 7, characterized in that: The step of performing multi-depth correlation mining on the target automobile finance feature representation and the associated automobile finance feature representation to mine semantic information having correlation with the associated automobile finance feature representation from the target automobile finance feature representation to obtain corresponding correlation feature representations includes: In the first depth correlation mining, semantic information having correlation with the associated automobile finance feature representation is mined from the target automobile finance feature representation to obtain the first depth correlation feature representation; In the correlation mining at other depths other than the first depth, a local masking operation is performed on the correlation feature representation of the previous depth to obtain a local masked feature representation of the current depth, and semantic information having correlation with the local masked feature representation of the current depth is mined from the target automobile finance feature representation to obtain a correlation feature representation of the current depth; After obtaining the correlation feature representation of the last depth, the correlation feature representations of each depth are fused to form the final correlation feature representation.

9. The artificial intelligence-based automobile finance solution matching method according to any one of claims 1 to 8, characterized in that: The automobile finance solution matching method based on artificial intelligence also includes: Acquire first automobile finance data corresponding to a first automobile finance user, and acquire second automobile finance data corresponding to a second automobile finance user associated with the first automobile finance user, wherein the first automobile finance data at least includes user identity information and user property information of the first automobile finance user, and the second automobile finance data at least includes user identity information and user property information of the second automobile finance user, and the first automobile finance data and the second automobile finance data are used as sample data for training a neural network; Using a feature extraction model included in a candidate financial solution matching network, feature extraction operations are performed on the first automobile financial data and the second automobile financial data respectively, and a first automobile financial feature representation corresponding to the first automobile financial data and a second automobile financial feature representation corresponding to the second automobile financial data are output, wherein the candidate financial solution matching network is a neural network; Using the financial solution matching model included in the candidate financial solution matching network, taking the second automobile financial feature representation as an auxiliary feature representation of the first automobile financial feature representation, and analyzing and outputting a financial solution matching probability distribution corresponding to the first automobile financial user; Based on the financial solution matching probability distribution corresponding to the first automobile finance user and the actually selected automobile finance solution corresponding to the first automobile finance user, a training error of the candidate financial solution matching network is analyzed, and based on the training error, a network parameter included in the candidate financial solution matching network is updated to form a target financial solution matching network corresponding to the candidate financial solution matching network, wherein the target financial solution matching network is used to analyze and output the target automobile finance solution corresponding to the target automobile finance user.

10. An automobile finance solution matching system based on artificial intelligence, characterized in that: include: Memory for storing computer programs; A processor connected to the memory is used to execute the computer program stored in the memory to implement the artificial intelligence-based automobile financial solution matching method described in any one of claims 1 to 9.

Images