Financial product recommendation method and device, equipment and medium

Abstract

The invention relates to the technical field of data analysis, and relates to a financial product recommendation method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining historical financial management information of a target customer, determining financial management label information of the target customer according to the historical financial management information, extracting candidate products matched with the financial management label information from a preset database, then calculating a matching degree score, a quality score and a satisfaction degree score of each candidate product respectively, and determining the matching degree score, the quality score and the satisfaction degree score of each candidate product according to the matching degree score, the quality score and the satisfaction degree score. And carrying out comprehensive operation on the three scores to obtain comprehensive analysis values, and finally pushing a preset number of candidate products with the comprehensive analysis values ranked in the front to the target customer. According to the method, the matching degree score, the quality score and the satisfaction score of the candidate product are respectively calculated, and the three scores are subjected to comprehensive operation, so that the score corresponding to the candidate product is accurately calculated, and the accuracy of financing product recommendation is improved.

Description

Technical Field

[0001] The present invention relates to the fields of data analysis technology, intelligent recommendation technology and financial technology, and in particular to a financial product recommendation method, device, equipment and medium. Background Art

[0002] With the development of the economy and the increase in national income, people's awareness of investment and financial management is also increasing, and the demand for various financial products is also increasing. As the financial institution that people come into contact with the most, banks need to help users find financial products that suit them. The existing financial product recommendation method is usually a one-to-one recommendation by bank employees to customers, which relies more on manual experience. And because of the limited access to customer information, bank employees generally only recommend more popular financial products. The products are too single and fixed, making the financial products recommended to customers inaccurate.

[0003] In order to solve the above problems, AI-based financial product recommendation methods are becoming more and more popular. However, existing AI-based financial product recommendations only consider the user's financial preferences, and do not combine the quality and evaluation of the financial products themselves, which affects the accuracy of the financial products recommended to customers. Therefore, how to improve the accuracy of financial product recommendations has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In view of the above, it is necessary to provide a financial product recommendation method to solve the technical problem that only the user's financial preferences are considered without combining the quality and evaluation of the financial product itself, thereby making the accuracy of financial product recommendations low.

[0005] To achieve the above object, the present invention provides a financial product recommendation method, which comprises:

[0006] Upon receiving a financial product recommendation request from a target customer, obtaining historical financial information of the target customer, and determining financial label information of the target customer based on the historical financial information, wherein the financial label information includes financial preference information;

[0007] Extracting candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data;

[0008] Calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm;

[0009] According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products with high comprehensive analysis value rankings are pushed to the target customers.

[0010] In addition, to achieve the above-mentioned purpose, the present invention also provides a financial product recommendation device, the device comprising:

[0011] Recommendation request receiving module: upon receiving a financial product recommendation request sent by a target customer, obtaining the target customer's historical financial information, and determining the target customer's financial label information according to the historical financial information, wherein the financial label information includes financial preference information;

[0012] Candidate product extraction module: extracts candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data;

[0013] Score calculation module: calculates the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculates the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculates the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm;

[0014] Candidate product recommendation module: According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products ranked high in the comprehensive analysis value are pushed to the target customers.

[0015] In addition, to achieve the above object, the present invention further provides an electronic device, the electronic device comprising:

[0016] a memory storing at least one computer program; and

[0017] The processor executes the program stored in the memory to implement the financial product recommendation method described above.

[0018] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored, and the at least one computer program is executed by a processor in an electronic device to implement the above-mentioned financial product recommendation method.

[0019] In the above technical solution provided by the present invention, by obtaining the historical financial information of the target customer, determining the financial label information of the target customer according to the historical financial information, extracting the candidate products matching the financial label information from the preset database, and then using the preset similarity algorithm to calculate the matching score of each candidate product with the financial preference information, using the pre-trained text quality scoring model to calculate the quality score of the product description text of each candidate product, according to the historical user operation data of each candidate product and using the preset user satisfaction calculation algorithm to calculate the satisfaction score of each candidate product, and then according to the preset score weight algorithm, the matching score, quality score and value score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and the preset number of candidate products with the top ranking of the comprehensive analysis value are pushed to the target customer. By respectively calculating the matching score, quality score and satisfaction score of the candidate product, and performing a comprehensive operation on the three scores, a more accurate comprehensive analysis value of each candidate product is obtained, and then the product to be recommended is selected according to the comprehensive analysis value, thereby improving the accuracy of financial product recommendation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0021] Figure 1 is a schematic diagram of an application environment of an intelligent question-answering processing method in one embodiment of the present invention;

[0022] Figure 2 A schematic diagram of a flow chart of a financial product recommendation method provided by an embodiment of the present invention;

[0023] Figure 3 A schematic diagram of the structure of a financial product recommendation device provided by an embodiment of the present invention;

[0024] Figure 4 A schematic diagram of the internal structure of an electronic device for implementing a financial product recommendation method provided by an embodiment of the present invention;

[0025] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0027] It should be noted that the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0028] The financial product recommendation method provided by the embodiment of the present invention can be applied in Figure 1 In an application environment, the client communicates with the server through a network. The server can receive the user's financial product recommendation request through the client, obtain the historical financial information of the target customer, and determine the financial label information of the target customer according to the historical financial information; use a preset similarity algorithm to calculate the matching score of each candidate product and the financial preference information, use a pre-trained text quality scoring model to calculate the quality score of the product description text of each candidate product, and calculate the satisfaction score of each candidate product according to the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm; according to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products with a high ranking of the comprehensive analysis value are pushed to the target customer. In the present invention, by respectively calculating the matching score, quality score and satisfaction score of the candidate product, and performing a comprehensive operation on the three scores, a more accurate comprehensive analysis value of each candidate product is obtained, and then the product to be recommended is selected according to the comprehensive analysis value, thereby improving the accuracy of financial product recommendation. The client may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices. The server may be implemented by an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0029] Reference Figure 2The flowchart of the method for recommending financial products provided by an embodiment of the present invention is shown in FIG. 1 . In the embodiment of the present invention, the method for recommending financial products includes the following steps S1-S4:

[0030] S1. Upon receiving a financial product recommendation request from a target customer, obtaining historical financial information of the target customer, and determining financial label information of the target customer based on the historical financial information, wherein the financial label information includes financial preference information.

[0031] In one embodiment, the financial label information further includes a risk tolerance level, and determining the financial label information of the target customer based on the historical financial information includes:

[0032] According to the historical financial management information, the historical financial management amount of the target customer is counted, and the risk tolerance level of the target customer is determined according to the historical financial management amount;

[0033] According to the historical financial information, the types of financial products purchased by the target customer and the corresponding purchase times are counted, and the financial preference information of the target customer is determined according to the types of financial products and the corresponding purchase times.

[0034] In this embodiment, the historical financial information refers to the financial products and related financial data purchased by the target customer in the past period of time, including historical financial product orders and corresponding historical financial amounts. By obtaining the historical financial information of the target customer, the financial product types and corresponding purchase times of the target customer can be counted from the historical financial information to determine the financial preference information of the target customer. The financial preference information includes a preference type label and a preference score corresponding to the preference type label. For example, the preference type label includes investing in the stock market, bonds and demand deposits, and the corresponding preference scores are 0.7, 0.2 and 0.1 respectively. The risk tolerance level is divided into the first level (high risk), the second level (medium risk) and the third level (low risk). The historical financial amount corresponding to each financial product order can also be obtained from the historical financial information, and the risk tolerance level of the target customer is determined according to the historical financial amount. For example, if the historical financial amount of the target customer is greater than or equal to 100,000 yuan, the target customer is identified as the first level; if the historical financial amount of the target customer is less than 100,000 yuan and greater than or equal to 50,000 yuan, the target customer is identified as the second level; if the historical financial amount of the target customer is less than 50,000 yuan, the target customer is identified as the third level. It is understandable that this embodiment only provides a classification of risk tolerance levels, and does not limit the risk tolerance levels. The risk tolerance levels can be classified according to actual conditions.

[0035] In another embodiment of the present invention, if the historical financial information of the target customer cannot be found, the risk tolerance level and financial preference information of the target customer can also be obtained by setting up a "customer risk tolerance assessment system" and a "customer preference information questionnaire".

[0036] In this embodiment, when the target customer needs to obtain financial product information through an electronic device, the financial product recommendation system pre-installed on the electronic device can be automatically triggered. Specifically, the financial product recommendation system can be automatically triggered through a website, or through an APP. After detecting the target customer's access information, the electronic device automatically triggers the "Customer Risk Tolerance Assessment System" and "Customer Preference Information Questionnaire" and displays them for the target customer to fill in the customer risk tolerance information and financial preference information, thereby obtaining the target customer's risk tolerance level and financial preference information.

[0037] In one embodiment, after the target customer completes filling in the financial preference information, the target customer's score for the financial product type may be further collected to obtain the preference score corresponding to each preference type label.

[0038] After the target customer completes the "Customer Risk Tolerance Assessment System" and the customer preference information questionnaire, click Submit. After detecting the target customer's submitted information, the electronic device determines the target customer's risk tolerance level based on the customer risk tolerance information submitted by the target customer, and stores the target customer's risk tolerance level and financial management preference information. When the target customer clicks again to trigger the preset financial management control in the financial management product recommendation system, the electronic device responds to the operation of triggering the financial management control and obtains the target customer's risk tolerance level and financial management preference information.

[0039] S2. Extract candidate products matching the financial management tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data.

[0040] In one embodiment, the candidate products correspond to product description texts and historical user operation data. The preset database is a pre-established large-scale corpus, including various types of financial products and some popular science content related to financial management. The preset database can be set by technical personnel during development, or adjusted by users during use, and the embodiment of the present invention is not limited to this. The popular science content about financial products in the corpus, in particular, can be used to establish a large-scale electronic text library through scientific sampling and processing. These text libraries can include various introductions, analyses, risk warnings, etc. about financial products, which are used in linguistic research, natural language processing and other fields. After obtaining the corpus, preprocessing is required, including removing punctuation, numbers, stop words, etc., to facilitate subsequent analysis and processing.

[0041] This embodiment matches corresponding candidate products from a preset database according to the target customer's financial label information. For example, a financial product with a preferred type label of money fund and a risk tolerance level of the second level is selected.

[0042] S3. Calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm.

[0043] In one embodiment, the calculating the matching score between each candidate product and the financial preference information by using a preset similarity algorithm includes:

[0044] According to a predetermined financial product keyword library, extracting a first keyword from a product description text corresponding to each of the candidate products, and extracting a second keyword from a preference type tag in the financial preference information;

[0045] Convert the first keyword into a vector to obtain a first word vector, and convert the second keyword into a vector to obtain a second word vector;

[0046] The matching score between the first word vector and the second word vector is calculated using a preset similarity algorithm.

[0047] In one embodiment, the predetermined financial product keyword library contains keywords corresponding to each type of financial product, such as bonds, stocks, funds and treasury bonds. The preset similarity algorithm can be a cosine similarity or a Euclidean distance algorithm. The first keyword is extracted from the product description text corresponding to each candidate product. For example, in the product description text of the candidate product whose financial product type is bonds, the first keyword "bond" is extracted; in the product description text of the candidate product whose financial product type is investment in the stock market, the first keyword "stock", "stock selection" and the like are extracted. The second keyword is extracted from the preference type label in the financial preference information, for example, the second keyword "bond" is extracted from the preference type label of "bond-type financial products", and the second keyword "stock" is extracted from the preference type label of "stock investment". It should be noted that the above is only a preferred embodiment of the present invention and is not limited thereto. The conversion of the first keyword into the first word vector and the conversion of the second keyword into the second word vector are usually implemented by word embedding or one-hot, etc. The method is a commonly used technical means in the field and will not be explained and illustrated in detail here.

[0048] This embodiment uses low-dimensional vectors to represent keywords, so that keywords with the same semantics are closer in the vector space, which can better describe the similarity between the preference type label and the product description text corresponding to the candidate product, thereby improving the accuracy and success rate of subsequent recommendations of financial products to users.

[0049] In one embodiment, the calculating the quality score of the product description text of each candidate product by using the pre-trained text quality score model includes:

[0050] Segmenting the product description text of the candidate product into a plurality of text sentences of equal length according to a predefined text length;

[0051] Calculating the cross entropy of the plurality of text sentences using the text quality score model, and using the cross entropy as the sub-quality score of the corresponding text sentence;

[0052] A preset algorithm is used to calculate a comprehensive quality value of the sub-quality scores of different text sentences, and the comprehensive quality value is used as the quality score of the product description text of each candidate product.

[0053] In this embodiment, the pre-trained text quality score model is an RNN model trained by text sequence samples of determined quality. The cross entropy of the plurality of text sentences is calculated by using the text quality score model, specifically:

[0054] For sentence S: "This is an example" can be processed into the following groups (x, y):

[0055] x1 = "Start this is", y1 = "This is a";

[0056] x2 = "This is a", y2 = "It is a";

[0057] x3 = "is a", y3 = "an example";

[0058] x4="an example", y4="an example".

[0059] In a specific application scenario, the quality of text sentences can be directly quantified by the cross-entropy loss of sequence prediction. The cross-entropy describes the distance between the actual output (probability) and the expected output (probability). The smaller the value of the cross-entropy, the closer the two probability distributions are. Take the above groups (x, y) as an example: calculate the cross-entropy of the loss function of the text sentences respectively. That is, x is the input data of the language model, and the expected output of the input data x is obtained through the language model. The quality score of this group (x, y) is determined by calculating the distance between the actual output y and the probability distribution of the expected output. The comprehensive quality score of the sentence S: "This is an example" is determined by calculating the mean of the scores of different groups (x, y). This embodiment helps customers better evaluate and understand the characteristics and risks of financial products by calculating the text quality of the financial products themselves, so as to make more informed investment decisions.

[0060] In one embodiment, the calculating the satisfaction score of each candidate product according to the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm includes:

[0061] Obtaining historical user operation data for each of the candidate products;

[0062] Calculate the information value scores of various types of data in the historical user operation data by using an information value algorithm, and use the data whose information value scores are greater than or equal to the information value threshold as reference data;

[0063] The satisfaction score of each candidate product is calculated based on the reference data using a preset user satisfaction calculation algorithm.

[0064] In one embodiment, the historical user operation data is data generated when the user browses or uses the candidate product, including user basic information, user behavior data, user interaction data, transaction data, user feedback and other information. The information value algorithm is used to determine the various types of data in the historical user operation data that have an impact on the satisfaction of the candidate product, for example, the user's evaluation, click, collection, forwarding and other information on the candidate product in the user interaction data. The confidence value algorithm determines the value of various types of data by calculating the value score. The information value threshold is a preset threshold, and the data with an information value parameter higher than the threshold is used as the reference data of the historical user operation data. Then, based on the preset user satisfaction calculation algorithm, the satisfaction score of each candidate product is calculated according to the reference data. The preset user satisfaction calculation algorithm includes calculating the RATER index. The RATER index is a comprehensive satisfaction index composed of five dimensions, namely reliability, assurance, tangibles, empathy and responsiveness. This embodiment can clearly know the satisfaction of historical users with the financial product by calculating the satisfaction score of the financial product, which is conducive to recommending financial products with high evaluation to customers.

[0065] This embodiment performs a comprehensive calculation on the quality score, matching score and satisfaction score of the candidate products, which helps to comprehensively evaluate the performance of the candidate products, thereby improving the accuracy of subsequent financial product recommendations to target customers.

[0066] S4. Perform weighted summation of the matching score, quality score and satisfaction score corresponding to each candidate product according to a preset score weight algorithm to obtain a comprehensive analysis value for each candidate product, and push a preset number of candidate products ranked high in the comprehensive analysis value to the target customers.

[0067] The step of pushing a preset number of candidate products with top comprehensive analysis values ​​to the target customers includes:

[0068] sorting the candidate products in descending order according to the corresponding comprehensive analysis values, selecting a preset number of candidate products ranked first as target candidate products, and pushing the target candidate products to the target customer; or

[0069] The candidate products are sorted in ascending order according to the corresponding comprehensive analysis values, a preset number of candidate products in the latter order are selected as target candidate products, and the target candidate products are pushed to the target customers.

[0070] In this embodiment, the candidate products are sorted in order of the size of the comprehensive analysis values ​​to obtain an analysis sorting table, and then the top X or bottom X (if from large to small, the top X are selected; if from small to large, the bottom X are selected) candidate products in the analysis sorting table are selected as target candidate products, where X is an integer greater than zero.

[0071] In another embodiment of the present invention, the step of pushing a preset number of candidate products with top comprehensive analysis values ​​to the target customer includes:

[0072] Screen out candidate products whose comprehensive analysis values ​​are greater than a preset threshold, and analyze whether the number of screened candidate products is greater than a preset number;

[0073] If the number is greater than a preset number, a preset number of candidate products with the highest comprehensive analysis values ​​are selected from the screened candidate products as target candidate products, and the target candidate products are pushed to the target customer;

[0074] If it is less than or equal to the preset number, all the screened candidate products are used as target candidate products, and the target candidate products are pushed to the target customers.

[0075] Specifically, the preset threshold is a standard threshold set in advance. Under normal circumstances, the number of candidate products with a comprehensive analysis value greater than the preset threshold cannot be determined, and there may be a large or small number. In order to recommend enough financial products to target customers, this embodiment will obtain a preset number of candidate products with a comprehensive analysis value greater than the preset threshold by setting. If the number of candidate products with a comprehensive analysis value greater than the preset threshold is greater than the preset number, then the preset number of candidate products with the comprehensive analysis value ranked first will be directly selected. If it is less than or equal to, then all the selected candidate products will be recommended to the target customer. In addition, in the case where the number of selected candidate products is less than the preset number, this embodiment can also adjust the preset threshold to lower the threshold, so as to select a sufficient number of candidate products.

[0076] In one embodiment, the product to be recommended to the target customer is usually presented to the user in the form of text, image, video, etc., so that the user can understand and choose the financial product.

[0077] In the above technical solution provided by the present invention, by obtaining the historical financial information of the target customer, determining the financial label information of the target customer according to the historical financial information, extracting the candidate products matching the financial label information from the preset database, and then using the preset similarity algorithm to calculate the matching score of each candidate product with the financial preference information, using the pre-trained text quality scoring model to calculate the quality score of the product description text of each candidate product, according to the historical user operation data of each candidate product and using the preset user satisfaction calculation algorithm to calculate the satisfaction score of each candidate product, and then according to the preset score weight algorithm, the matching score, quality score and value score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and the preset number of candidate products with the top ranking of the comprehensive analysis value are pushed to the target customer. By respectively calculating the matching score, quality score and satisfaction score of the candidate product, and performing a comprehensive operation on the three scores, a more accurate comprehensive analysis value of each candidate product is obtained, and then the product to be recommended is selected according to the comprehensive analysis value, thereby improving the accuracy of financial product recommendation.

[0078] like Figure 3 FIG. 1 is a schematic diagram of the structure of the financial product recommendation device of the present invention.

[0079] The financial product recommendation device 100 of the present invention can be installed in an electronic device. According to the functions to be implemented, the financial product recommendation device can include a recommendation request receiving module 101, a candidate product extraction module 102, a score calculation module 103, and a candidate product recommendation module 104. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by a processor of an electronic device and can complete fixed functions, and is stored in the memory of the electronic device.

[0080] In this embodiment, the functions of each module / unit are as follows:

[0081] The recommendation request receiving module 101 is used to obtain the historical financial information of the target customer when receiving the financial product recommendation request sent by the target customer, and determine the financial label information of the target customer according to the historical financial information, wherein the financial label information includes financial preference information;

[0082] A candidate product extraction module 102 is used to extract candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data;

[0083] The score calculation module 103 is used to calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm;

[0084] The candidate product recommendation module 104 is used to perform weighted summation of the matching score, quality score and satisfaction score corresponding to each candidate product according to a preset score weight algorithm to obtain a comprehensive analysis value for each candidate product, and push a preset number of candidate products ranked high in the comprehensive analysis value to the target customers.

[0085] Furthermore, the financial label information also includes a risk tolerance level. The recommendation request receiving module 101 is specifically used to:

[0086] According to the historical financial management information, the historical financial management amount of the target customer is counted, and the risk tolerance level of the target customer is determined according to the historical financial management amount;

[0087] According to the historical financial information, the types of financial products purchased by the target customer and the corresponding purchase times are counted, and the financial preference information of the target customer is determined according to the types of financial products and the corresponding purchase times.

[0088] The score calculation module 103 is specifically used for:

[0089] According to a predetermined financial product keyword library, extracting a first keyword from a product description text corresponding to each of the candidate products, and extracting a second keyword from a preference type tag in the financial preference information;

[0090] Convert the first keyword into a vector to obtain a first word vector, and convert the second keyword into a vector to obtain a second word vector;

[0091] The matching score between the first word vector and the second word vector is calculated using a preset similarity algorithm.

[0092] The score calculation module 103 is also used for:

[0093] Segmenting the product description text of the candidate product into a plurality of text sentences of equal length according to a predefined text length;

[0094] Calculating the cross entropy of the plurality of text sentences using the text quality score model, and using the cross entropy as the sub-quality score of the corresponding text sentence;

[0095] A preset algorithm is used to calculate a comprehensive quality value of the sub-quality scores of different text sentences, and the comprehensive quality value is used as the quality score of the product description text of each candidate product.

[0096] The score calculation module 103 is also used for:

[0097] Obtaining historical user operation data for each of the candidate products;

[0098] Calculate the information value scores of various types of data in the historical user operation data by using an information value algorithm, and use the data whose information value scores are greater than or equal to the information value threshold as reference data;

[0099] The satisfaction score of each candidate product is calculated based on the reference data using a preset user satisfaction calculation algorithm.

[0100] The candidate product recommendation module 104 is specifically used for:

[0101] sorting the candidate products in descending order according to the corresponding comprehensive analysis values, selecting a preset number of candidate products ranked first as target candidate products, and pushing the target candidate products to the target customer; or

[0102] The candidate products are sorted in ascending order according to the corresponding comprehensive analysis values, a preset number of candidate products in the latter order are selected as target candidate products, and the target candidate products are pushed to the target customers.

[0103] The candidate product recommendation module 104 is further used for:

[0104] Screen out candidate products whose comprehensive analysis values ​​are greater than a preset threshold, and analyze whether the number of screened candidate products is greater than a preset number;

[0105] If the number is greater than a preset number, a preset number of candidate products with the highest comprehensive analysis values ​​are selected from the screened candidate products as target candidate products, and the target candidate products are pushed to the target customer;

[0106] If it is less than or equal to the preset number, all the screened candidate products are used as target candidate products, and the target candidate products are pushed to the target customers.

[0107] The financial product recommendation device 100 provided by the present invention calculates the matching score, quality score and satisfaction score of the candidate products respectively, and performs comprehensive operation on the three scores to obtain a more accurate comprehensive analysis value for each candidate product, and then selects the product to be recommended based on the comprehensive analysis value, thereby improving the accuracy of financial product recommendations.

[0108] like Figure 4FIG. 1 is a schematic diagram of the internal structure of an electronic device for implementing the financial product recommendation method of the present invention.

[0109] The electronic device may include a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a financial product recommendation program.

[0110] In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules (such as financial product recommendations, etc.) stored in the memory 11, and calls data stored in the memory 11 to execute various functions of the electronic device and process data.

[0111] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 11 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 can be used not only to store application software and various types of data installed in the electronic device, such as the code of a financial product recommendation program, but also to temporarily store various types of data that have been output or are to be output.

[0112] The communication bus 12 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The communication bus 12 is configured to realize connection and communication between the memory 11 and at least one processor 10, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0113] Figure 4 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 4 The structure shown does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0114] For example, although not shown, the electronic device may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure classification circuits, power converters or inverters, power status indicators, etc. The electronic device may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0115] Optionally, the communication interface 13 may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices.

[0116] Optionally, the communication interface 13 may also include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0117] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, device, article or method including the element.

[0118] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. Through the description of the above implementation modes, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course, by hardware, but in many cases the former is a better implementation mode. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0119] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

[0120] The financial product recommendation program stored in the memory 11 in the electronic device is a combination of multiple computer programs. When running in the processor 10, it can achieve:

[0121] Upon receiving a financial product recommendation request from a target customer, obtaining historical financial information of the target customer, and determining financial label information of the target customer based on the historical financial information, wherein the financial label information includes financial preference information;

[0122] Extracting candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data;

[0123] Calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm;

[0124] According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products with high comprehensive analysis value rankings are pushed to the target customers.

[0125] Specifically, the specific implementation method of the processor 10 for the above computer program can refer to Figure 1 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0126] Furthermore, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable medium can be non-volatile or volatile. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM).

[0127] An embodiment of the present invention may also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program may implement:

[0128] Upon receiving a financial product recommendation request from a target customer, obtaining historical financial information of the target customer, and determining financial label information of the target customer based on the historical financial information, wherein the financial label information includes financial preference information;

[0129] Extracting candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data;

[0130] Calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm;

[0131] According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products with high comprehensive analysis value rankings are pushed to the target customers.

[0132] Furthermore, the computer-usable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.

[0133] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0134] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0135] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0136] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0137] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0138] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0139] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0140] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The second and other words are used to indicate names, but not to indicate any particular order.

[0141] Finally, it should be noted that if software tools or components other than those of the Company appear in the embodiments of the present application, they are only used for illustration and do not represent actual use. The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for recommending financial products, characterized in that: The method comprises: Upon receiving a financial product recommendation request from a target customer, obtaining historical financial information of the target customer, and determining financial label information of the target customer based on the historical financial information, wherein the financial label information includes financial preference information; Extracting candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data; Calculate the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculate the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculate the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm; According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products with high comprehensive analysis value rankings are pushed to the target customers.

2. The financial product recommendation method according to claim 1, characterized in that: The financial label information also includes a risk tolerance level. The financial label information of the target customer is determined based on the historical financial information, including: According to the historical financial management information, the historical financial management amount of the target customer is counted, and the risk tolerance level of the target customer is determined according to the historical financial management amount; According to the historical financial information, the types of financial products purchased by the target customer and the corresponding purchase times are counted, and the financial preference information of the target customer is determined according to the types of financial products and the corresponding purchase times.

3. The financial product recommendation method according to claim 1, characterized in that: The using a preset similarity algorithm to calculate the matching score between each candidate product and the financial preference information includes: According to a predetermined financial product keyword library, extracting a first keyword from a product description text corresponding to each of the candidate products, and extracting a second keyword from a preference type tag in the financial preference information; Convert the first keyword into a vector to obtain a first word vector, and convert the second keyword into a vector to obtain a second word vector; The matching score between the first word vector and the second word vector is calculated using a preset similarity algorithm.

4. The financial product recommendation method according to claim 1, characterized in that: The using a pre-trained text quality score model to calculate the quality score of the product description text of each candidate product includes: Segmenting the product description text of the candidate product into a plurality of text sentences of equal length according to a predefined text length; Calculating the cross entropy of the plurality of text sentences using the text quality score model, and using the cross entropy as the sub-quality score of the corresponding text sentence; A preset algorithm is used to calculate a comprehensive quality value of the sub-quality scores of different text sentences, and the comprehensive quality value is used as the quality score of the product description text of each candidate product.

5. The financial product recommendation method according to claim 1, characterized in that: The step of calculating the satisfaction score of each candidate product according to the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm includes: Obtaining historical user operation data for each of the candidate products; Calculate the information value scores of various types of data in the historical user operation data by using an information value algorithm, and use the data whose information value scores are greater than or equal to the information value threshold as reference data; The satisfaction score of each candidate product is calculated based on the reference data using a preset user satisfaction calculation algorithm.

6. The financial product recommendation method according to claim 1, characterized in that: The step of pushing a preset number of candidate products with top comprehensive analysis values ​​to the target customers includes: sorting the candidate products in descending order according to the corresponding comprehensive analysis values, selecting a preset number of candidate products ranked first as target candidate products, and pushing the target candidate products to the target customer; or The candidate products are sorted in ascending order according to the corresponding comprehensive analysis values, a preset number of candidate products in the latter order are selected as target candidate products, and the target candidate products are pushed to the target customers.

7. The financial product recommendation method according to claim 1, characterized in that: The step of pushing a preset number of candidate products with top comprehensive analysis values ​​to the target customers includes: Screen out candidate products whose comprehensive analysis values ​​are greater than a preset threshold, and analyze whether the number of screened candidate products is greater than a preset number; If the number is greater than a preset number, a preset number of candidate products with the highest comprehensive analysis values ​​are selected from the screened candidate products as target candidate products, and the target candidate products are pushed to the target customer; If it is less than or equal to the preset number, all the screened candidate products are used as target candidate products, and the target candidate products are pushed to the target customers.

8. A financial product recommendation device, characterized in that: The device comprises: Recommendation request receiving module: upon receiving a financial product recommendation request sent by a target customer, obtaining the target customer's historical financial information, and determining the target customer's financial label information according to the historical financial information, wherein the financial label information includes financial preference information; Candidate product extraction module: extracts candidate products matching the financial tag information from a preset database, wherein the candidate products correspond to product description texts and historical user operation data; Score calculation module: calculates the matching score between each candidate product and the financial preference information using a preset similarity algorithm, calculates the quality score of the product description text of each candidate product using a pre-trained text quality scoring model, and calculates the satisfaction score of each candidate product based on the historical user operation data of each candidate product and using a preset user satisfaction calculation algorithm; Candidate product recommendation module: According to the preset score weight algorithm, the matching score, quality score and satisfaction score corresponding to each candidate product are weighted and summed to obtain the comprehensive analysis value of each candidate product, and a preset number of candidate products ranked high in the comprehensive analysis value are pushed to the target customers.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the financial product recommendation method as described in any one of claims 1 to 7.

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

Images