A recommendation method, device, equipment and program product for credit investigation products

Abstract

The specification provides a recommendation method, device, equipment and program product for credit investigation products. The method comprises: in the case that the business demand description information does not contain the customer detail data of the to-be-evaluated customer group, obtaining business data reflecting user business attributes from a public channel, and analyzing the business data to determine the customer group characteristics of the target service object corresponding to the target business carried out by the business initiator; inputting the customer group characteristics into a preset mapping model to determine the target credit investigation distribution characteristics corresponding to the target service object through the mapping model; wherein the credit investigation distribution characteristics are used to represent the numerical distribution of the credit investigation data of the corresponding customer group in at least one credit investigation dimension; and determining a credit investigation product meeting the business demand description information and recommending it to the business initiator according to the target credit investigation distribution characteristics.

Description

Technical Field

[0001] This specification relates to one or more embodiments in the field of credit reporting technology, and more particularly to a recommendation method, apparatus, device, and program product for credit reporting products. Background Technology

[0002] With the continuous development of financial lending, financial institutions are increasingly relying on external credit reporting products in pre-loan approval, risk assessment, and customer segmentation. To meet compliance and business decision-making needs, credit reporting agencies typically develop credit reporting products that financial institutions can access, based on multi-source information and combined with credit reporting rules, risk control strategies, and related service functions. How to recommend suitable credit reporting products based on the business needs of financial institutions has become a crucial technical challenge in the credit reporting service field.

[0003] However, current recommendations for credit reporting products typically rely on customer samples provided by the business initiator—that is, detailed data on the customer group to be evaluated. This data is analyzed to determine the credit reporting product that matches the business needs. But when business initiators actually state their needs, they often only provide vague descriptions, making the existing sample-based recommendation method difficult to implement. This makes it impossible to accurately recommend credit reporting products that meet the actual business needs of the initiator when detailed customer data is lacking. Summary of the Invention

[0004] In view of the above, one or more embodiments of this specification provide the following technical solutions: According to a first aspect of one or more embodiments of this specification, a recommendation method for credit reporting products is proposed, comprising: Obtain the business requirement description information provided by the business initiator; If the business requirement description information does not contain customer details data of the customer group to be evaluated, business data reflecting the corresponding business attributes of the business initiator is obtained from public channels, and the business data is parsed to determine the customer group characteristics of the target service objects corresponding to the target business carried out by the business initiator. The customer group characteristics are input into a preset mapping model to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension; Based on the target credit information distribution characteristics, credit information products that match the business requirement description information are identified and recommended to the business initiator.

[0005] According to a second aspect of one or more embodiments of this specification, a recommendation apparatus for credit reporting products is provided, comprising: The acquisition module is used to acquire business requirement description information provided by the business initiator. The parsing module is used to obtain business data reflecting the corresponding business attributes of the business initiator from public channels when the business requirement description information does not contain customer detail data of the customer group to be evaluated, and to parse the business data to determine the customer group characteristics of the target service object corresponding to the target business carried out by the business initiator. The mapping module is used to input the customer group characteristics into a preset mapping model, so as to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension; The recommendation module is used to determine credit reporting products that match the business requirement description information based on the target credit reporting distribution characteristics and recommend them to the business initiator.

[0006] According to a third aspect of one or more embodiments of this specification, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor implements the steps of the method described above by executing the executable instructions.

[0007] According to a fourth aspect of one or more embodiments of this specification, a computer-readable storage medium is provided that stores computer instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0008] According to a fifth aspect of one or more embodiments of this specification, a computer program product is provided, comprising a computer program / instructions that, when executed by a processor, implement the steps of the method described above.

[0009] As can be seen from the above embodiments, when the business demand description information provided by the business initiator does not include detailed customer data of the customer group, this specification determines the customer group characteristics involved in the target business carried out by the business initiator by collecting and parsing the business data of the business initiator, and further converts the customer group characteristics into target credit distribution characteristics based on a preset mapping model, thereby determining and recommending credit products that meet the business needs of the business initiator based on the target credit distribution characteristics.

[0010] Compared to existing technologies that rely on customers' detailed customer data for testing and analysis, this specification can complete the characterization of business needs and matching of credit reporting products without requiring the business initiator to provide detailed customer data. This effectively solves the problem of inaccurate credit reporting product recommendations in scenarios with zero or missing samples. Furthermore, by transforming business semantic information into quantifiable credit reporting distribution features, this application helps improve the accuracy of demand understanding and the targeting of product recommendations, thereby increasing the efficiency of credit reporting product recommendations and the degree to which the recommendations match actual business needs. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the architecture of a recommendation service system provided in an exemplary embodiment; Figure 2 This is a flowchart illustrating a recommendation method for credit reporting products provided in an exemplary embodiment; Figure 3 This is an exemplary embodiment of a flowchart illustrating the overall determination process of a credit reporting product. Figure 4 This is a schematic diagram of the structure of a device provided in an exemplary embodiment; Figure 5 This is a block diagram of an apparatus for a recommendation method for credit reporting products, provided in an exemplary embodiment. Detailed Implementation

[0012] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0013] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of related data shall comply with the relevant laws, regulations and standards of the relevant regions, and corresponding operation portals shall be provided for users to choose to authorize or refuse.

[0014] In credit reporting service scenarios, financial institutions typically raise credit reporting product consultation requests based on their own business plans. These requests may include descriptive requirements such as "credit business to expand into lower-tier customer groups," "business to identify high-liquidity customer risk groups," or "business for small-amount, short-term loan access." These requests usually reflect business objectives or directions rather than structured customer data characteristics. During the actual consultation phase, the initiating party often cannot simultaneously provide detailed customer data for the customer group, or may be unable to provide any testable detailed customer data due to compliance, privacy, or cold start considerations for new business.

[0015] While existing technologies offer credit reporting product recommendations based on sample testing, historical backtesting, or human experience, several shortcomings remain in practical applications. First, current solutions typically rely on clients' detailed customer data, determining product suitability through offline testing, feature analysis, or performance evaluation on samples. When the business initiator fails to provide samples, these solutions lack basic input, preventing the recommendation process from starting. Second, the business needs descriptions provided by business initiators often possess strong semantic attributes, such as describing the industry, region, credit limit range, or target customer group, lacking quantitative features directly corresponding to credit reporting dimensions. Existing keyword or tag matching solutions struggle to accurately understand the business essence, leading to low recommendation accuracy. Third, although credit reporting agencies typically accumulate vast amounts of historical business test data, test reports, and records of product usage, this data is often unstructured, isolated, or fragmented across projects, making it difficult to reuse historical experience, especially to migrate it to new business scenarios without samples. Fourth, when recommendations are made based on human business or expert experience, the results are easily influenced by subjective experience, lack unified and quantitative judgment criteria, and suffer from problems such as insufficient recommendation stability, insufficient interpretability, and low efficiency.

[0016] The aforementioned deficiencies collectively make it difficult for existing technologies to accurately recommend credit reporting products that meet the actual business needs of the business initiator when the business demand description information does not include detailed customer data of the customer group.

[0017] Based on this, this specification provides a recommendation method for credit reporting products, which can still complete the entire recommendation process from understanding business needs and inferring customer characteristics to matching credit reporting products even in scenarios with no samples or missing samples, thereby improving the accuracy, automation, and adaptability of credit reporting product recommendations to actual business needs.

[0018] To facilitate understanding of the technical solution of this application, some key concepts involved in this application are explained below: Business initiator: Institutional entities that have needs for accessing, consulting, selecting, or using credit reporting products, such as banks, consumer finance companies, trust companies, microfinance companies, and other financial institutions engaged in credit business; of course, the business initiator can also refer to other entities that have needs for risk assessment, customer identification, or business decision support, but this specification does not make specific limitations on this.

[0019] Credit reporting products refer to products developed by credit reporting agencies based on multi-source information, combined with credit reporting rules, risk control strategies, assessment logic, or related service functions, and available for use by business initiators in their business decision-making processes. Credit reporting products can be single-data products or outputs such as scores, tags, rule results, risk assessment results, or other products with business decision-making support capabilities, processed from multiple data dimensions. In this application, credit reporting products are not limited to the aggregation results of raw data; they can also include functional or rule-based products for business scenarios such as customer access, risk identification, customer segmentation, credit limit assessment, and anti-fraud identification. Different credit reporting products can correspond to different applicable customer groups, business scenarios, and performance characteristics.

[0020] Business requirement description information: This refers to the descriptive information provided by the business initiator to express its own business needs. This information can be at least one of the following: natural language text, form fields, inquiry questions, business specification documents, and chat logs.

[0021] Business data refers to a collection of information that reflects the business operations of the initiator and can be legally obtained. Business data may include at least one of the following: publicly disclosed business introductions, publicly available web page content, tender notices, promotional materials, product introductions, publicly available interviews, and publicly available reports.

[0022] Target business: This refers to the business that, after identification, statistical analysis, and refinement from the business data of the business initiator, can characterize the core business features of the business initiator. A business initiator may engage in multiple business directions, and the target business is used to reflect the business content that is most relevant to the current business needs, most comprehensive, or most representative of the business focus of the business initiator.

[0023] Customer characteristics: These refer to the feature information used to characterize the customer groups involved in the target business. Customer characteristics may include one or more of the following: occupational attributes, geographical attributes, income stability, consumption activity, mobility characteristics, borrowing behavior characteristics, device usage characteristics, and identity stability characteristics. Customer characteristics can be described using tags, structured feature fields, or vectorized expressions.

[0024] Credit distribution characteristics refer to the features used to characterize the numerical distribution of credit data for a corresponding customer group across at least one credit dimension. Credit dimensions may include at least one of the following: multiple borrowing-related dimensions, device activity-related dimensions, online presence duration-related dimensions, identity stability-related dimensions, consumption behavior-related dimensions, and credit performance-related dimensions. Credit distribution characteristics can be represented using forms such as mean, variance, quantiles, interval proportions, discrete distribution, probability density, and vector distribution.

[0025] A mapping model is a pre-defined model used to establish the correspondence between customer characteristics and credit distribution characteristics. This mapping model can be a model trained based on supervised learning, a hybrid model combining rules and models, or a neural network model, tree model, graph model, or vector mapping model. The purpose of the mapping model is to transform the business-level understanding of customer groups into quantifiable credit distribution characteristics, providing a unified input for subsequent credit product recommendations.

[0026] The technical solutions described in the embodiments of this specification will be explained in detail below with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the architecture of a recommendation service system provided in an exemplary embodiment. Figure 1 As shown, the system may include a server 11, a network 12, and several electronic devices, such as a personal computer (PC) 13, a mobile phone 14, etc.

[0027] Server 11 can be a physical server containing an independent host, or it can be a virtual server hosted in a host cluster. During operation, server 11 can run server-side programs for a specific application to implement the relevant functions of that application. For example, when server 11 runs a recommendation service program, it can function as a corresponding recommendation service platform.

[0028] PC13 and mobile phone14 are just some of the types of electronic devices that the business initiator can use. In reality, the business initiator can obviously also use electronic devices such as tablets, laptops, PDAs (Personal Digital Assistants), wearable devices (such as smart glasses, smartwatches, etc.), etc., and one or more embodiments in this specification do not limit this. During operation, the electronic device can run a client-side program of an application to implement the relevant functions of that application. For example, when the electronic device runs a recommendation service program, it can act as a client for that recommendation service. The aforementioned recommendation service client application can be launched and run on the electronic device. This client-side program can be a native application installed on the electronic device, or it can be a mini-program, quick app, or other similar form. Of course, when using web technologies such as HTML5, the relevant functions can be implemented through a page displayed by a browser. This browser can be a standalone browser application or a browser module embedded in some applications.

[0029] As for the network 12 that enables interaction between electronic devices such as PC13 and mobile phone 14 and server 11, communication can be achieved using either wired or wireless networks, depending on the communication methods supported by the respective electronic devices. This specification does not impose any restrictions on this. For example, PC13 can support both wired and wireless communication, so it can use either wired or wireless networks as needed. Mobile phone 14 typically only supports wireless communication, so it can use a wireless network for communication.

[0030] In this specification, the executing entity for implementing the recommended methods for credit reporting products can be a designated device such as a server deployed in the credit reporting system. Of course, it can also be a device such as a mobile phone or computer and a client installed on these devices. For ease of description, the following will use a server as the executing entity to explain the recommended methods for credit reporting products provided in this specification.

[0031] Figure 2 This is a flowchart illustrating a recommendation method for credit reporting products, provided in an exemplary embodiment, including the following steps: S200: Obtain the business requirement description information provided by the business initiator.

[0032] The server can first obtain the business requirement description information provided by the business initiator. This description can be submitted by the business initiator through methods such as page input boxes, API call parameters, message sessions, requirement forms, and business consultation documents. The business requirement description can be a single sentence or multiple paragraphs of text. For example, the business initiator could input "I want to select a credit reporting product for my lending business," "I want to integrate with a credit reporting product," or "I want to match a suitable credit reporting product." The server can perform basic preprocessing on the business requirement description information, such as word segmentation, noise reduction, entity recognition, keyword extraction, and text vectorization, to facilitate subsequent business parsing.

[0033] After obtaining the business requirement description information, the server can further determine whether the business requirement description information contains customer detail data of the customer group. When it is detected that the business requirement description information does not contain customer detail data of the customer group, the process proceeds to S202.

[0034] In practical applications, servers can determine the existence of customer detail data through field integrity checks. For example, if the business requirement description information does not contain structured or unstructured sample content such as customer identifier sets, sample feature files, sample backtracking files, or customer detail records, it is determined that the customer detail data does not contain the customer group.

[0035] In addition, the server can also determine whether the current input belongs to a zero-sample consultation scenario based on a text classification model. If the classification result indicates that the current input only contains business descriptions and not customer details, then a subsequent recommendation process based on business data is triggered.

[0036] S202: If the business requirement description information does not contain customer details data of the customer group to be evaluated, obtain business data reflecting the corresponding business attributes of the business initiator from public channels, and parse the business data to determine the customer group characteristics of the target service objects corresponding to the target business carried out by the business initiator.

[0037] After confirming that the business requirement description information does not include a sample of the customer group to be evaluated (i.e., customer detail data), the server can obtain business data reflecting the corresponding business attributes of the business initiator from public channels. These business attributes may include the target business conducted by the business initiator, the business scenario of the target business, and its target service recipients. The target service recipients may refer to the set of direct or potential customer groups targeted by the target business.

[0038] Business data can originate from multiple channels. For example, a server can scrape information such as the business introduction, product description, press releases, and publicly available marketing materials from publicly available web pages; it can obtain information such as the business plan, business focus, and target customer group description from publicly available bidding materials, reports, or interviews; and, with authorization, it can supplement its understanding of the business initiator's existing business direction by combining information such as historical contract records, historical consultation records, or business interface call logs. The above sources are merely examples, and the specific sources of business data are not limited.

[0039] After obtaining business data, the server can parse the data to determine the customer characteristics of the target customer group involved in the business initiated by the business initiator. First, the server can identify the business scenario information contained in the business data. Specifically, the server can perform natural language processing on the text content of the business data to extract business scenario information that can represent the business type, customer base, purpose, marketing target, geographic coverage, product structure, etc. For example, from "small-amount revolving business for flexible employment groups," the server can extract business scenario information such as "flexible employment groups" and "small-amount revolving business"; from "serving county-level business operators and self-employed individuals," the server can extract business scenario information such as "county-level business operators" and "self-employed individuals."

[0040] Secondly, the server performs statistical analysis on business scenario information to determine the target business scenario that represents the core business characteristics of the business initiator. This statistical analysis is not limited to simple counting; it can also include frequency statistics, text clustering, topic extraction, semantic similarity calculation, keyword weight analysis, and time window weighted analysis. For example, the server can count the frequency of different business scenario information in multiple sets of business data and select the business scenario with the highest frequency as the target business scenario; it can also assign higher weights to business scenarios appearing in recent announcements, core product introduction pages, or key business sections based on preset business importance rules, and then select the business scenario with the highest comprehensive score as the target business scenario; or it can cluster multiple business scenario information and select the scenario corresponding to the cluster center as the target business scenario. Through this process, target business scenarios that represent the core business characteristics of the business initiator can be extracted from multi-source, scattered, and noisy business data.

[0041] Secondly, based on the target business scenario, the server determines the customer characteristics of the target service recipients corresponding to the target business. Specifically, the server can map the target business scenario to a set of customer characteristics. For example, when the target business scenario points to "small cash loan business targeting highly mobile blue-collar customers," the server can determine that the corresponding customer characteristics include: occupational attributes leaning towards blue-collar, geographical distribution leaning towards lower-tier markets, higher income volatility, weaker social security stability, higher nighttime activity, and moderate to high equipment replacement frequency. When the target business scenario points to "business lending business targeting self-employed individuals," the server can determine that the customer characteristics include: obvious business attributes, strong demand for working capital, high transaction activity, moderate to high loan amount demand, and the importance of stable operation and repayment ability.

[0042] In some embodiments, customer group characteristics can be determined through a rule engine. For example, a "target business scenario - customer group characteristics" mapping rule base can be pre-established. Once the target business scenario is identified, the corresponding customer group characteristics can be directly returned through the rule base. In other embodiments, customer group characteristics can also be determined through large-scale model inference, classification models, multi-label prediction models, or knowledge base retrieval. For example, the target business scenario can be input into a trained feature inference model, which outputs a set of customer group characteristic labels and their confidence scores; then, the final customer group characteristics are selected based on the confidence score threshold. Alternatively, the server can simultaneously employ rule mapping and model inference, and fuse the results of both to improve the stability of customer group characteristic determination.

[0043] In one example scenario, the business initiator only provides a description of its business needs, stating that it "hopes to use the cash loan business in lower-tier markets," without providing any customer samples. The server can identify from the business data that the initiator has a long-term focus on consumer credit in lower-tier markets. Statistical analysis determines that "small-amount cash loan business targeting highly mobile customers in lower-tier cities" is the target business scenario. Further inferences suggest that the target customer characteristics include: geographically concentrated in lower-tier cities, moderate to low income stability, a high probability of not having stable housing provident funds or social security, active mobile terminal usage, and frequent short-term funding needs.

[0044] Of course, in addition to publicly available data, the server can also obtain internal data from the business initiator, such as business solution documents, historical cooperation records, historical product access records, business configuration parameters, API call records, internal operational analysis data, or anonymized statistical reports. When the business initiator has established a cooperative relationship with a credit reporting agency, or agrees to provide additional business information in the current consultation process, the server can send a data acquisition request to the business initiator's system or the internal business management system associated with the business initiator. After obtaining authorization from the business initiator, the server can obtain the corresponding internal data, and then determine the aforementioned customer characteristics from the internal data or the combination of external and internal data.

[0045] In this specification, various intelligent agents can be pre-configured to perform specialized processing and collaborative analysis of information related to the business needs of the business initiator. These intelligent agents may include information gathering agents, intent understanding agents, and feature reasoning agents, among others.

[0046] The server can perform processing related to step S200 and subsequent business data collection through an information collection agent, used to retrieve and aggregate publicly available information and authorized internal information related to the business initiator; an intent understanding agent performs semantic analysis on the collected unstructured text to identify the business initiator's true business intent, business focus, and corresponding business scenario; and a feature reasoning agent, based on the identified business scenario, infers and completes the customer characteristics of the target service object under that business scenario, thereby generating feature representations for subsequent mapping. Through the collaborative processing of these multiple agents, the accuracy and completeness of business requirement analysis, target business identification, and customer characteristic determination can be improved even when the business initiator does not provide detailed customer data for the customer group to be evaluated.

[0047] It should be noted that if the description information includes business scenario information, then there is no need to obtain public or internal data.

[0048] S204: Input the customer group characteristics into a preset mapping model to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension.

[0049] After determining the characteristics of the customer group, the server can input them into a preset mapping model to determine the target credit distribution characteristics of the target service object.

[0050] The core of this step lies in further transforming the customer perception at the business semantic level into a statistical distribution expression in the credit data space, thereby opening up the mapping channel between "business language" and "credit language".

[0051] The mapping model can be pre-built based on historical training data. This historical training data can include multiple historical business sample pairs, each of which includes at least the business-side customer group characteristics and the corresponding credit distribution characteristics of that customer group. By learning from a large number of historical sample pairs, the mapping model can grasp the correspondence between different customer group characteristics and credit distribution characteristics. For example, for a customer group characterized by "highly mobile blue-collar workers," the mapping model can learn the statistical distribution characteristics that they typically exhibit in credit dimensions such as mobile phone number network duration, loan application frequency, active time distribution, device stability, and repayment fluctuations.

[0052] The mapping model can be implemented in various ways, including: The mapping model can be a supervised learning model, trained by taking customer group characteristics as input and credit distribution characteristics as output; The mapping model can be a deep neural network model, which encodes customer group features through embedding layers and outputs distribution parameters for multiple credit dimensions; The mapping model can also be a hybrid model, for example, first determining the distribution of some credit dimensions with relatively strong determinism through rules or knowledge bases, and then predicting the distribution of the remaining dimensions through machine learning models; ... For different implementation methods, as long as the determination of the target credit distribution characteristics from customer group characteristics can be achieved, this specification does not make specific limitations on this.

[0053] In this specification, credit distribution features are used to characterize the overall numerical distribution of credit data for a corresponding customer group across at least one credit dimension. There can be multiple ways to express the target credit distribution features: The target credit distribution characteristics can be described using continuous numerical statistics, such as the mean, standard deviation, skewness, and percentage of quantile intervals for a certain credit dimension. The target credit distribution characteristics can be described using discrete interval distribution, such as "X% of mobile phone numbers have been online for 12 to 24 months" and "Y% of multiple loans have been taken out 3 to 5 times". The target credit distribution features can also be expressed in a vectorized manner, that is, the statistical information of multiple credit dimensions is encoded into a distribution vector, which is then used for similarity matching and product retrieval. ... For example, after the server inputs customer characteristics such as "high mobility, lower-tier cities, no stable social security, and high nighttime activity" into the mapping model, the credit distribution characteristics output by the mapping model can be: The duration of mobile phone number usage is mainly concentrated in the range of 12 to 24 months, the average number of multiple loans is at a medium to high level, the proportion of active devices at night is higher than that of the general customer group, the proportion of missing data related to stable social security is relatively high, and the volatility of payment behavior is relatively high.

[0054] It should be noted that the target credit distribution feature output by the mapping model can be a single candidate result or multiple candidate distribution results. When multiple candidate distribution results are output, the server can select the target credit distribution feature from the multiple candidate distribution results based on confidence level, degree of matching with business requirement description information, or historical consistency score. Alternatively, multiple candidate distribution results can be weighted and fused to obtain the final target credit distribution feature.

[0055] Furthermore, if the description information of the business initiator includes a customer group, the server can directly jump from step S200 to S204, that is, directly determine the target credit distribution characteristics based on the customer detail data contained in the description information.

[0056] S206: Based on the target credit information distribution characteristics, determine the credit information products that match the business requirement description information and recommend them to the business initiator.

[0057] After determining the target credit information distribution characteristics, the server can identify credit information products that meet the business needs of the business initiator based on the target credit information distribution characteristics, and recommend the credit information product to the business initiator.

[0058] In practical applications, servers can pre-establish a database of compatibility relationships between credit reporting products and credit reporting distribution characteristics. Each credit reporting product can be associated with one or more compatible distribution characteristic ranges or applicable customer group distribution patterns. After obtaining the target credit reporting distribution characteristics, the server can perform matching in this database to determine the credit reporting product that matches the target credit reporting distribution characteristics.

[0059] For example, all credit reporting products with similarity higher than a preset threshold can be selected based on a similarity threshold; the top K credit reporting products with the highest similarity to the target credit reporting distribution characteristics can be directly selected based on Top-K sorting; or the matching degree of different credit reporting dimensions can be weighted and summed based on multi-dimensional weighted scoring, and the credit reporting product with the highest comprehensive score can be selected as the recommendation result.

[0060] In a preferred embodiment provided in this specification, the server can construct a product recommendation knowledge graph by combining historical business data, and complete the determination of credit reporting products based on the product recommendation knowledge graph.

[0061] Specifically, the server can acquire historical business data from the business initiator's historical business processes and extract various business elements contained within this data. These business elements include: historical business scenarios, historical credit reporting distribution characteristics, historical credit reporting products, and historical efficiency indicators. Then, the server constructs a product recommendation knowledge graph, using each business element as a node and the relationships between these elements as edges between nodes. This product recommendation knowledge graph can be used to accumulate experience regarding the correlation between historical business scenarios and the effectiveness of credit reporting products, enabling currently unsampled business scenarios to leverage historical experience for migration recommendations.

[0062] In one embodiment, the server can identify credit distribution feature nodes in the product recommendation knowledge graph that match the target credit distribution feature. This "matching" can be achieved in various ways, such as calculating the similarity between the target credit distribution feature and historical credit distribution feature nodes based on Euclidean distance, cosine similarity, Mahalanobis distance, or distribution similarity coefficient; identifying nodes with similarity scores above a threshold; or selecting the top few credit distribution feature nodes based on similarity ranking as the matching results. The server can then determine candidate credit products based on the credit product nodes associated with the credit distribution feature node, and select the credit product from the candidate products that meets the actual business needs of the business initiator. Although the node name includes "candidate credit product," the actual recommendation output still corresponds to the credit product recommended to the business initiator.

[0063] In another embodiment, the server can construct a target business subgraph based on the target business scenario and target credit distribution characteristics corresponding to the target business; then, it can extract historical business subgraphs that match the target business subgraph from the product recommendation knowledge graph; finally, based on the credit product nodes contained in the historical business subgraphs, it can determine the credit product that meets the actual business needs of the business initiator. The subgraph matching method here can include structural similarity matching, joint node attribute matching, path matching, or graph embedding vector matching. Compared to single-point matching based solely on credit distribution characteristic nodes, subgraph matching considers the joint relationship between the target business scenario and the target credit distribution characteristics, which helps improve the accuracy of credit product determination.

[0064] It should be noted that the aforementioned energy efficiency indicators can refer to indicator information used to characterize the actual effectiveness of credit reporting products under corresponding business scenarios or historical credit reporting distribution characteristics. These energy efficiency indicators can be at least one of the following: discrimination index, ranking ability index, stability index, and hit rate index.

[0065] For example, the discrimination index can be the Kolmogorov-Smirnov Statistic (KS), which is used to characterize the ability of a credit reporting product to distinguish customers of different risk categories. Different credit reporting products may have different associated efficiency indicators; of course, the same credit reporting product may also be associated with multiple different efficiency indicators.

[0066] It should be further added that the matching methods based on credit reporting distribution feature nodes and target business subgraphs mentioned above can be further constrained by the focus of attention reflected in the business requirement description information, so as to further improve the matching degree between the credit reporting product determination results and the actual business needs of the business initiator.

[0067] Specifically, in the process of identifying credit reporting products that meet the actual business needs of the business initiator in the knowledge graph, the server can use the business requirement description information as a basis, and further screen and determine the candidate credit reporting products based on the business requirement description information and the target credit reporting distribution characteristics.

[0068] When the business requirement description includes energy efficiency indicators of interest to the business initiator, during the node matching process based on the target credit distribution characteristics, for each matched credit distribution characteristic node, the server can determine the credit product associated with the target energy efficiency indicator from among the multiple credit product nodes associated with that credit distribution characteristic node, and use them as candidate credit products. Then, the server determines the target credit product from the candidate credit products.

[0069] Similarly, when performing subgraph matching, the server can construct a target business subgraph based on the target business scenario, target credit distribution characteristics, and target efficiency indicators contained in the business requirement description information. Then, it extracts historical business subgraphs from the product recommendation knowledge graph that match the target business subgraph and meet the target efficiency indicator requirements. Based on the credit product nodes associated with these historical business subgraphs, it determines the credit products that meet the target efficiency indicator requirements. Compared to subgraph matching based solely on the target business scenario and target credit distribution characteristics, this approach further incorporates the target efficiency indicators that the business initiator focuses on into the subgraph construction and matching process. This helps to filter out credit products that perform better in the specified performance dimension from historical business subgraphs, improving the relevance of the recommendation results to the target performance indicators.

[0070] Furthermore, the server can identify candidate credit reporting products within the product recommendation knowledge graph and determine the associated energy efficiency indicators for each candidate product. Based on each candidate credit reporting product and its corresponding energy efficiency indicator, it predicts the credit reporting capability assessment value for each candidate product. Then, based on the credit reporting capability assessment value, it selects at least one target credit reporting product from the candidate products that meets the actual business needs of the business initiator. The credit reporting capability assessment value here reflects the expected performance of the candidate credit reporting product under the current target credit reporting distribution characteristics. For example, it can comprehensively consider factors such as historical energy efficiency indicator values, the closeness of the current target distribution to historical distributions, the matching degree of the product's applicable scenarios, and product stability to output the credit reporting capability assessment value for each candidate credit reporting product.

[0071] There are also multiple ways to predict creditworthiness assessment scores: The server can employ rule-based calculation methods. For example, historical energy efficiency indicators can be used as the base score, and the similarity between the target credit distribution characteristics and historical credit distribution characteristics can be used as a correction coefficient. A weighted formula can then be used to calculate the creditworthiness assessment value. The server can use machine learning prediction models, taking candidate credit reporting product characteristics, historical energy efficiency indicators, target credit reporting distribution characteristics, etc. as inputs, and the model outputs a credit reporting capability assessment value. The server can use graph neural networks or path reasoning models to directly score candidate credit products on the product recommendation knowledge graph; ... Furthermore, the server can not only compare individual candidate credit reporting products, but also combine them to obtain at least one credit reporting product combination, and determine the credit reporting capability assessment value corresponding to each combination. Then, the candidate credit reporting product and the combination with the highest corresponding credit reporting capability assessment value are selected as the target credit reporting product that meets the actual business needs of the business initiator. Here, the target credit reporting product can be understood as the final recommendation result, which can be either a single credit reporting product or a combination of credit reporting products.

[0072] Product portfolios can be constructed in pairs, according to preset complementary rules, by dimensional coverage, or by heuristic search methods. For example, if one credit reporting product focuses on identifying identity stability and another focuses on identifying shared debt risk, the server can combine the two and evaluate their overall suitability and expected effect on the current target credit reporting distribution characteristics.

[0073] In one example scenario, the server retrieves three candidate credit reporting products from the credit reporting graph based on the target credit reporting distribution characteristics. The first candidate credit reporting product has a high historical efficiency index, but is mainly suitable for stable customer groups; the second candidate credit reporting product is suitable for highly mobile customer groups; and the third candidate credit reporting product is suitable for identifying multiple risk. The server predicts that the credit reporting capability assessment value of the second candidate credit reporting product in the current business scenario is higher than that of the first candidate credit reporting product. Furthermore, the combined credit reporting capability assessment value of the second and third candidate credit reporting products is higher than that of any single product. Therefore, the server can determine this combination of credit reporting products as the final recommendation result and recommend it to the business initiator.

[0074] When recommending products to the business initiator, the server can output recommendation explanation information in addition to the final recommended credit scoring product. This explanation information may include the target business scenario, a summary of customer characteristics, a summary of the target credit scoring distribution characteristics, the basis for the recommendation, the ranking results of candidate products, and the estimated effect range. This helps improve the interpretability of the recommendation results and facilitates business decision-making for the business initiator. It should be noted that the recommendation explanation information is optional output and does not constitute an additional limitation on the scope of protection of the claims.

[0075] To facilitate understanding, the above method can be further illustrated with the following complete example: Suppose a business initiator submits a business requirement description stating "a credit reporting product suitable for newly launched customer groups," without providing any customer details. The server first retrieves this business requirement description and identifies the current scenario as a no-sample consultation.

[0076] Subsequently, the server collects business data from the business initiator, including publicly available business introductions on the official website, key business directions mentioned in public interviews, and product layout information in publicly available market materials. From this business data, the server identifies multiple business scenario information, such as "targeting county-level customers," "targeting highly mobile blue-collar workers," and "targeting small-amount cash turnover scenarios," and through statistical analysis, determines "small-amount cash turnover business targeting highly mobile lower-tier customers" as the target business scenario.

[0077] Next, the server determines customer characteristics based on the target business scenario, such as a high proportion of lower-tier markets, low stable social security coverage, high nighttime activity, and strong short-term turnover needs.

[0078] The server then inputs the customer group characteristics into the mapping model to obtain the target credit distribution characteristics, such as the mobile phone number's online duration being concentrated between 12 and 24 months, the number of multiple loans being in the medium to high range, and the device activity fluctuating significantly.

[0079] Finally, the server retrieves candidate credit reporting products from the product recommendation knowledge graph based on the target credit reporting distribution characteristics, and predicts the credit reporting capability assessment value of each candidate credit reporting product and its combination by combining historical energy efficiency indicators. From this, the server selects the target credit reporting product with the highest assessment value and recommends it to the business initiator. Thus, even without samples, the server can still complete a relatively accurate credit reporting product recommendation.

[0080] Furthermore, this specification also provides an overall flowchart for determining credit reporting products, such as... Figure 3 As shown.

[0081] Figure 3 This is an exemplary embodiment of a flowchart for determining the overall process of a credit reporting product.

[0082] The process can be divided into three parts: the perception and reasoning layer, the knowledge and mapping layer, and the decision and interpretation layer. The perception and reasoning layer performs requirement understanding and feature inference processing based on the business requirement description information provided by the business initiator. The server can collect, analyze, and infer relevant business data and authorized internal data from the business initiator through intelligence gathering agents, intent understanding agents, and feature inference agents to identify the true business intent of the business initiator, determine the target business scenario corresponding to the target business, and infer the customer characteristics of the customer group involved in the target business scenario. The knowledge and mapping layer is used for... The aforementioned customer characteristics are converted into quantifiable target credit information distribution characteristics, and then related retrieval and graph reasoning are performed based on a product recommendation knowledge graph. The product recommendation knowledge graph can include nodes such as historical business scenarios, historical credit information distribution characteristics, credit information products, and historical energy efficiency indicators to support similar scenario matching, subgraph matching, and candidate product screening. The decision and interpretation layer is used to determine the target credit information products that meet the actual business needs of the business initiator based on the target credit information distribution characteristics and graph reasoning results, and generate corresponding recommendation results and explanation information, thereby completing the credit information product recommendation process in scenarios where customer group details are lacking.

[0083] As can be seen from the above method, this specification, even when the business requirement description information provided by the business initiator does not include detailed customer data of the customer group, determines the customer characteristics of the target business group by collecting and analyzing the business data of the business initiator, and determines the target credit reporting distribution characteristics based on a preset mapping model. Then, based on the target credit reporting distribution characteristics, it determines the credit reporting product that meets the business initiator's business requirements. Therefore, this specification can complete credit reporting product recommendations in scenarios with zero or missing samples, effectively solving the technical problem that existing technologies struggle to accurately recommend credit reporting products without samples.

[0084] Furthermore, this specification realizes the transformation from business requirement description to credit distribution characteristics, which helps to narrow the gap between business semantics and product selection, improve the understanding of the actual business needs of the business initiator, and enhance the relevance and adaptability of the recommendation results.

[0085] In addition, this manual can also combine product recommendation knowledge graphs to conduct correlation analysis on historical business scenarios, historical credit reporting distribution characteristics, credit reporting products and historical efficiency indicators, thereby realizing the effective reuse of historical experience and further improving the accuracy and efficiency of credit reporting product recommendations.

[0086] Figure 4 This is a schematic structural diagram of a device provided in an exemplary embodiment. For example... Figure 4 As shown, device 400 mainly consists of a communication interface 402, a service initiator interface 404, a processor 406, and a data storage 408. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 410. Communication interface 402 enables device 400 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, communication interface 402 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, communication interface 402 can be a wired interface such as Ethernet, Token Ring, or USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or wide area wireless interface (e.g., WiMAX or LTE). Of course, communication interface 402 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. Communication interface 402 may also include multiple physical communication interfaces, such as Wi-Fi, Bluetooth, and wide area wireless interfaces.

[0087] The service initiator interface 404 includes receiving service initiator input and providing output to the service initiator. Therefore, the service initiator interface 404 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. The service initiator interface 404 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, the service initiator interface 404 may include software, circuitry, or other forms of logic capable of transmitting and receiving data from external service initiator input / output devices. Additionally or alternatively, device 400 may support remote access from other devices via communication interface 402 or another physical interface (not shown). The service initiator interface 404 may be configured to receive service initiator input, the position and movement of which may be indicated by indicators or cursors described herein. The business initiator interface 404 can also be configured as a display device for rendering or displaying text fragments.

[0088] Processor 406 may contain one or more general-purpose processors and / or special-purpose processors.

[0089] Data storage 408 may include one or more volatile and / or non-volatile storage components and may be integrated wholly or partially with processor 406. Data storage 408 may include removable and non-removable components.

[0090] Processor 406 is capable of executing program instructions 418 (e.g., compiled or uncompiled program logic and / or machine code) stored in data storage 408 to perform the various functions described herein. Data storage 408 may comprise a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 400, enable device 400 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Processor 406 executing program instructions 418 may result in processor 406 using data 412.

[0091] For example, program instructions 418 may include an operating system 422 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 400 and one or more applications 420 (e.g., a browser, social application, or game application). Similarly, data 412 may include operating system data 416 and application data 414. Operating system data 416 is primarily accessible to the operating system 422, while application data 414 is primarily accessible to one or more applications 420. Application data 414 may reside in a file system visible or hidden from the service initiator of device 400.

[0092] Application 420 can communicate with operating system 422 through one or more application programming interfaces (APIs). These APIs help application 420 read and / or write application data 414, transmit or receive information via communication interface 402, receive or display information on service initiator interface 404, etc.

[0093] In some terminology, application 420 may be simply referred to as "app". Furthermore, application 420 can be downloaded to device 400 through one or more online app stores or app markets. However, applications can also be installed on device 400 in other ways, such as through a web browser or a physical interface on device 400 (e.g., a USB port).

[0094] Please refer to Figure 5 Recommendation devices for credit reporting products can be applied to, for example... Figure 4 The device shown implements the technical solution described in this specification. The recommended device for credit reporting products may include: Module 500 is used to obtain the business requirement description information provided by the business initiator. The parsing module 502 is used to obtain business data reflecting the corresponding business attributes of the business initiator from public channels when the business requirement description information does not contain customer detail data of the customer group to be evaluated, and to parse the business data to determine the customer group characteristics of the target service object corresponding to the target business carried out by the business initiator. The mapping module 504 is used to input the customer group characteristics into a preset mapping model, so as to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension; The recommendation module 506 is used to determine, based on the target credit distribution characteristics, a credit product that meets the actual business needs of the business initiator and recommend it to the business initiator.

[0095] Optionally, the parsing module 502 is specifically used to: identify the business scenario information contained in the business data; perform statistical analysis on the business scenario information to determine the target business scenario used to characterize the core business characteristics of the business initiator; and determine the customer group characteristics based on the target business scenario.

[0096] Optionally, the acquisition module 500 is further configured to acquire historical business data during the execution of historical business by the business initiator, and extract various business elements contained in the historical business data, wherein the business elements include: historical business scenarios, historical credit distribution characteristics, historical credit products, and historical energy efficiency indicators; and construct a product recommendation knowledge graph with each business element as a node and the relationship between each business element as the edge between the nodes. The recommendation module 506 is specifically used to determine, based on the target credit distribution characteristics, a credit product that meets the actual business needs of the business initiator in the product recommendation knowledge graph.

[0097] Optionally, the recommendation module 506 is specifically used to: determine credit distribution feature nodes that match the target credit distribution features in the product recommendation knowledge graph; determine candidate credit products based on the credit product nodes associated with the credit distribution feature nodes; and determine credit products that meet the actual business needs of the business initiator from the candidate credit products.

[0098] Optionally, the recommendation module 506 is specifically used to: construct a target business subgraph based on the target business scenario corresponding to the target business and the target credit distribution characteristics; extract historical business subgraphs that match the target business subgraph from the product recommendation knowledge graph; and determine credit products that meet the actual business needs of the business initiator based on the credit product nodes contained in the historical business subgraphs.

[0099] Optionally, the recommendation module 506 is specifically configured to: determine each candidate credit reporting product in the product recommendation knowledge graph based on the target credit reporting distribution characteristics, and determine the energy efficiency indicators associated with each candidate credit reporting product; predict the credit reporting capability assessment value corresponding to each candidate credit reporting product based on each candidate credit reporting product and its corresponding energy efficiency indicators; the credit reporting capability assessment value is used to reflect the expected effect of the candidate credit reporting product under the target credit reporting distribution characteristics; and select at least one target credit reporting product that meets the actual business needs of the business initiator from among the candidate credit reporting products based on the credit reporting capability assessment value.

[0100] Optionally, the recommendation module 506 is specifically used to: predict the credit capability assessment value corresponding to each candidate credit reporting product; combine each candidate credit reporting product to obtain at least one credit reporting product combination, and determine the credit capability assessment value corresponding to each credit reporting product combination; and determine the highest credit capability assessment value among each candidate credit reporting product and each credit reporting product combination as the target credit reporting product that meets the actual business needs of the business initiator.

[0101] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0102] Based on the same concept as the methods described above, this specification also provides an electronic device, including: a processor; a memory for storing processor-executable instructions; wherein the processor performs the steps of the method as described in any of the above embodiments by executing the executable instructions.

[0103] Based on the same concept as the methods described above, this specification also provides a computer-readable storage medium having computer instructions stored thereon that, when executed by a processor, implement the steps of the methods as described in any of the above embodiments.

[0104] Based on the same concept as the methods described above, this specification also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the methods as described in any of the above embodiments.

[0105] What those skilled in the art will understand is: In this specification, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded.

[0106] In this specification, “a,” “an,” and “the” do not specifically refer to the singular, but may also include the plural.

[0107] In this specification, ordinal numbers such as "first," "second," etc., do not necessarily indicate order; they are often used to distinguish between objects. For example, "first server" and "second server" usually refer to two servers. To differentiate between these two servers, they are described as "first server" and "second server." Of course, sometimes these two servers may be the same server.

[0108] In this specification, unless explicitly stated otherwise, "receiving and sending data" does not necessarily mean direct receiving and sending; it can also mean indirect receiving and sending. For example, A receiving data sent by B can be understood as A directly receiving the data sent by B, or it can be understood as A indirectly receiving the data sent by B through other entities such as C. Similarly, B sending data to A can be understood as B sending the data directly to A, or it can be understood as B indirectly sending the data to A through other entities such as C. Here, C can be one entity, or it can be two or more entities.

[0109] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0110] This specification uses specific terms to describe embodiments thereof. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of those different embodiments or examples, without contradiction.

[0111] Although one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is only one of many possible execution orders and does not represent the only execution order. Therefore, when the claims involve method steps, any changes or adjustments to the order of such steps, or the parallelism between steps, are also within the scope of protection of the claims.

Claims

1. A recommendation method for credit reporting products, comprising: Obtain the business requirement description information provided by the business initiator; If the business requirement description information does not contain customer details data of the customer group to be evaluated, business data reflecting the corresponding business attributes of the business initiator is obtained from public channels, and the business data is parsed to determine the customer group characteristics of the target service objects corresponding to the target business carried out by the business initiator. The customer group characteristics are input into a preset mapping model to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension; Based on the target credit information distribution characteristics, credit information products that match the business requirement description information are identified and recommended to the business initiator.

2. The method as described in claim 1, wherein determining the customer characteristics of the target service recipients corresponding to the target business conducted by the business initiator specifically includes: Identify the business scenario information contained in the business data; Statistical analysis is performed on the business scenario information to determine the target business scenario used to characterize the core business features of the business initiator; Based on the target business scenario, the characteristics of the customer group are determined.

3. The method of claim 1, further comprising: Obtain historical business data from the process of the business initiator executing historical business, and extract the various business elements contained in the historical business data, wherein the business elements include: historical business scenarios, historical credit distribution characteristics, historical credit products, and historical energy efficiency indicators; A product recommendation knowledge graph is constructed using each business element as a node and the relationships between each business element as edges between nodes. The credit reporting products that meet the aforementioned business requirements description information include: Based on the target credit information distribution characteristics, credit information products that match the business requirement description information are determined in the product recommendation knowledge graph.

4. The method as described in claim 3, wherein determining credit reporting products that match the business requirement description information in the product recommendation knowledge graph, specifically includes: In the product recommendation knowledge graph, identify credit distribution feature nodes that match the target credit distribution features; Candidate credit reporting products are determined based on the credit reporting product nodes associated with the credit reporting distribution feature nodes; From the candidate credit reporting products, determine the credit reporting products that match the business requirement description information.

5. The method as described in claim 3, wherein determining credit reporting products that match the business requirement description information in the product recommendation knowledge graph, specifically includes: Based on the target business scenario corresponding to the target business and the target credit distribution characteristics, a target business sub-graph is constructed; Extract historical business subgraphs that match the target business subgraph from the product recommendation knowledge graph; Based on the credit reporting product nodes contained in the historical business sub-graph, determine the credit reporting products that meet the business requirement description information.

6. The method as described in claim 3, wherein determining credit reporting products that match the business requirement description information in the product recommendation knowledge graph, specifically includes: Based on the target credit distribution characteristics, each candidate credit product is identified in the product recommendation knowledge graph, and the energy efficiency index associated with each candidate credit product is determined. Based on each candidate credit reporting product and its corresponding energy efficiency index, the credit reporting capability assessment value corresponding to each candidate credit reporting product is predicted; the credit reporting capability assessment value is used to reflect the expected effect of the candidate credit reporting product under the target credit reporting distribution characteristics. Based on the credit reporting capability assessment value, at least one target credit reporting product that meets the business requirement description information is selected from among the candidate credit reporting products.

7. The method as described in claim 6, wherein, based on the credit reporting capability assessment value, at least one target credit reporting product that meets the actual business needs of the business initiator is selected from among the candidate credit reporting products, specifically including: Predict the credit reporting capability assessment value corresponding to each candidate credit reporting product; The candidate credit reporting products are combined to obtain at least one credit reporting product combination, and the credit reporting capability assessment value corresponding to each credit reporting product combination is determined. The credit reporting product with the highest corresponding credit reporting capability assessment value among all candidate credit reporting products and combinations of credit reporting products is determined as the target credit reporting product that meets the business requirement description information.

8. A recommendation device for credit reporting products, comprising: The acquisition module is used to acquire business requirement description information provided by the business initiator. The parsing module is used to obtain business data reflecting the corresponding business attributes of the business initiator from public channels when the business requirement description information does not contain customer detail data of the customer group to be evaluated, and to parse the business data to determine the customer group characteristics of the target service object corresponding to the target business carried out by the business initiator. The mapping module is used to input the customer group characteristics into a preset mapping model, so as to determine the target credit distribution characteristics corresponding to the target service object through the mapping model; wherein, the credit distribution characteristics are used to characterize the numerical distribution of the credit data of the corresponding customer group in at least one credit dimension; The recommendation module is used to determine credit reporting products that match the business requirement description information based on the target credit reporting distribution characteristics and recommend them to the business initiator.

9. An electronic device comprising: processor; A memory for storing processor-executable instructions; wherein the processor implements the steps of the method as described in any one of claims 1-7 by executing the executable instructions.

10. A computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the method as claimed in any one of claims 1-7.

Images