Method for identifying financing expiration event based on public data to form business opportunity and pushing business opportunity

Abstract

The invention discloses a method for identifying financing expiration events based on public data to form business opportunities and pushing the business opportunities. The pushing method comprises the steps that S1, multivariate heterogeneous data collection and structured modeling are carried out; s2, constructing a financing expiration event business opportunity model; s3, calculation and sorting output of business opportunity recommendation scores; and S4, constructing a business opportunity intelligent pushing mechanism. The integration and structured modeling of multi-source heterogeneous data are realized, and the comprehensiveness and accuracy of financing event identification are improved. And constructing a business opportunity recommendation model fused with multi-dimensional features, and realizing dynamic evaluation and intelligent sorting of business opportunities.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of commercial bank public business, and in particular to a method for identifying financing expiration events based on public data to form business opportunities and pushing. BACKGROUND

[0002] The prior art has many deficiencies in identifying financing expiration events and pushing business opportunities, which are embodied in the following aspects:

[0003] Through the bank's own credit management system (such as the core system, CRM), the loan contract expiration date and repayment plan of the inventory customers are tracked, and an automatic early warning mechanism is set. Some banks introduce AI models to predict the probability of loan renewal, such as modeling based on historical repayment behavior, account flow volatility and other characteristics. But this can only monitor the bank's credit customers, and cannot identify the financing expiration events of customers in other banks (such as cross-bank loans, trust financing), nor can it identify the financing expiration events of non-bank customers.

[0004] The traditional business opportunity mining process highly depends on manual investigation, regular visits and other methods of customer managers, which is low in efficiency and difficult to scale. Especially in the scene of complex financing structure of small and medium-sized enterprises and non-transparent account, the manual identification error rate is high, and it is difficult to form an effective business opportunity closed loop.

[0005] The marketing business opportunities pushed to branches and customer managers lack pertinence, and there is a common problem of "full push, no differential marketing", and there is a lack of intelligent matching mechanism based on customer relationship graph, which leads to the fact that business opportunities cannot be accurately allocated to the most suitable operating units or customer managers, resulting in resource waste and low business opportunity conversion rate. SUMMARY

[0006] In view of the above problems, the present application is proposed in order to provide a method for identifying financing expiration events based on public data to form business opportunities and pushing, which overcomes the above problems or at least partially solves the above problems.

[0007] According to one aspect of the present application, there is provided a method for identifying financing expiration events based on public data to form business opportunities and pushing, which comprises:

[0008] Step S1: multi-element heterogeneous data acquisition and structured modeling;

[0009] Step S2: constructing a financing expiration event business opportunity model;

[0010] Step S3: business opportunity recommendation calculation and sorting output;

[0011] Step S4: constructing a business opportunity intelligent pushing mechanism.

[0012] Optionally, the step S1: multi-element heterogeneous data acquisition and structured modeling specifically comprises:

[0013] Collecting enterprise-related data from multiple public data sources;

[0014] Data extraction and cleaning: using natural language processing (NLP) and information extraction technology to process unstructured text and extract key fields;

[0015] Building a financing expiration event database: structuring all extracted financing event data and building a financing expiration time database Each record is represented as:

[0016]

[0017] Wherein, E j is an entity; is the financing type; is the expiration date; is the financing amount; is the status.

[0018] Optionally, the collecting enterprise-related data from multiple public data sources specifically includes:

[0019] Business data: registered capital, legal person information, shareholder structure;

[0020] Judicial data: executed information, credit record, legal proceedings;

[0021] Financial data: annual report data, including balance sheet, income statement, cash flow statement;

[0022] Public opinion data: news, announcements, social media information;

[0023] Illegal data and penalty data: tax violations, administrative penalties;

[0024] Financing data: bank loans, trust financing, bond financing, financing leasing, accounts receivable.

[0025] Optionally, the key fields specifically include: financing amount, expiration date, financing type.

[0026] Optionally, the step S2: building a financing expiration event business opportunity model specifically includes:

[0027] Based on the financing expiration event database Building a financing expiration event business opportunity model, including: identifying financing expiration event business opportunities that meet the requirements, and using enterprise self-coefficient, distance coefficient, and business opportunity merging coefficient as model coefficients to form a business opportunity recommendation score.

[0028] Optionally, the based on the financing expiration event database Building a financing expiration event business opportunity model specifically includes:

[0029] Business opportunity identification criteria:

[0030] For database Each record f in j A business opportunity is considered to be one that meets the following conditions:

[0031] The due date is later than the current date and the balance is greater than 0. Data without a balance field defaults to a balance greater than 0.

[0032] Filter out companies that are currently overdue, have poor credit records, or have abnormal business operations according to Dazhihui (a business information platform).

[0033] Businesses must be located in the city where the bank's branch is located;

[0034]

[0035] Business Opportunity Recommendation Score Model Construction: Defining Business Opportunity Recommendation Score

[0036] in: The opportunity merging coefficient indicates whether the company has simultaneously captured other business opportunities (such as existing customers or high-quality customers);

[0037]

[0038] Where K represents the number of business opportunity types; I represents the number of business opportunity types. k : Whether the k-th type of business opportunity has been hit; λ k δ: Weighting coefficients for various superior levels; δ: Merging coefficient amplification factor;

[0039] Distance coefficient, representing the distance between the business and the bank branch;

[0040]

[0041] The company's own coefficient is composed of the following weighted sub-items:

[0042]

[0043] in: Enterprise nature coefficient; Enterprise size coefficient; Enterprise qualification coefficient; Industry coefficient; ω1+ω2+ω3+ω4=1: weights of each dimension;

[0044] Business opportunity ranking mechanism: All business opportunities are sorted in descending order of recommendation score. If the recommendation scores are the same, they are ranked according to...

[0045] Based on the financing maturity event opportunity model, a recommendation score for financing maturity event opportunities is determined, and a priority list of opportunities is generated, including:

[0046] Coefficient assignment and calculation: A combination of expert evaluation and historical data training is used to assign values ​​to each coefficient in the model, and dynamic optimization is performed through machine learning models.

[0047] Business opportunity screening mechanism: setting a recommendation score threshold R th Retain business opportunities with a referral score higher than this threshold: Qualified = {f j |R j >R th};

[0048] Output Business Opportunity List: Outputs a filtered list of priority business opportunities. Where k≤n, sorted by recommendation score.

[0049] Optionally, step S4: constructing a business opportunity intelligent push mechanism specifically includes:

[0050] Based on the business opportunity recommendation and distribution mechanism for financing maturity events, business opportunities are pushed to relevant banks through automatic or manual allocation methods.

[0051] Optionally, the business opportunity recommendation and push mechanism based on financing maturity events to reach business opportunities specifically includes:

[0052] Push method selection includes:

[0053] Automatic allocation: Target customers are associated with existing customers. Based on the customer relationship graph, they are matched and assigned to the organization of the existing customer with the highest correlation.

[0054] Manual allocation: Business opportunities are distributed layer by layer from the bank's head office, branches, and sub-branches;

[0055] The customer relationship matching algorithm includes: constructing a correlation matrix M between existing customers and target companies. ij ,in:

[0056]

[0057] Where: i is the existing customer ID; j is the target company ID; r ijk ω represents the degree of association of the k-th type of relation; k represents the weight of each type of relation; K represents the total number of relation types.

[0058] Select the business units of existing customers with the highest relevance to push business opportunities:

[0059]

[0060] The push execution and feedback mechanism includes: recording the marketing response status and conversion rate data after the push is sent, which is used for model iteration and optimization.

[0061] This invention provides a method for identifying and pushing business opportunities based on publicly available data related to financing maturity events. The method includes: Step S1: collecting and structurally modeling multi-source heterogeneous data; Step S2: constructing a business opportunity model for financing maturity events; Step S3: calculating and ranking the business opportunity recommendation score; Step S4: constructing an intelligent business opportunity push mechanism. This method integrates multi-source heterogeneous data and performs structured modeling, improving the comprehensiveness and accuracy of financing event identification. It also constructs a business opportunity recommendation model that integrates multi-dimensional features, enabling dynamic evaluation and intelligent ranking of business opportunities.

[0062] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0063] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 The flowchart illustrates a method for identifying and promoting business opportunities based on publicly available data in an embodiment of the present invention. Detailed Implementation

[0065] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0066] The terms "comprising" and "having," and any variations thereof, in the specification, embodiments, claims, and drawings of this invention are intended to cover non-exclusive inclusion, such as including a series of steps or units.

[0067] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0068] like Figure 1As shown, a method for identifying and pushing business opportunities based on publicly available data regarding financing maturity events includes:

[0069] Step S1: Collect and structure the heterogeneous data from multiple sources;

[0070] Step S2: Construct a business opportunity model for financing maturity events;

[0071] Step S3: Calculate and sort the business opportunity recommendations and output them;

[0072] Step S4: Build an intelligent business opportunity push mechanism.

[0073] Step 1: Multi-source heterogeneous data acquisition and structured modeling

[0074] (1) This step collects relevant enterprise data from multiple public data sources, including:

[0075] Business registration data: registered capital, legal representative information, shareholder structure;

[0076] Judicial data: information on those subject to enforcement, records of dishonesty, and legal proceedings;

[0077] Financial data: Annual report data, including balance sheet, income statement, and cash flow statement;

[0078] Public opinion data: news, announcements, and social media information;

[0079] Data on violations and penalties: tax violations, administrative penalties;

[0080] Financing data: bank loans, trust financing, bond financing, financial leasing, and accounts receivable.

[0081] (2) Data extraction and cleaning

[0082] Natural Language Processing (NLP) and information extraction techniques are used to process unstructured text and extract key fields such as "financing amount", "maturity date" and "financing type".

[0083] Let the text data set be D = {d1, d2, ..., dn}. n}, each text d i After information extraction, the structured triplet T is output. i = (E, A, V), where: E is the entity (such as company name); A is the attribute (such as financing type, maturity date); and V is the attribute value.

[0084] (3) Construct a database of financing maturity events

[0085] All extracted financing event data will be structured and stored to build a financing maturity time database.

[0086] Each record is represented as follows:

[0087]

[0088] Where: E j Entity (e.g., company name); Financing type (such as bank loans, bonds, etc.); Expiry date; Amount of financing; Status (whether it is overdue, whether it is in default, whether there is negative information).

[0089] Step Two: Construct a business opportunity model for financing maturity events

[0090] Based on the database of financing maturity events Construct a business opportunity model for financing maturity events, including: identifying eligible financing maturity event business opportunities, and using the company's own coefficient, distance coefficient, and business opportunity merging coefficient as model coefficients to form a business opportunity recommendation score.

[0091] (1) Business opportunity identification conditions

[0092] For database Each record f in j A business opportunity is considered to be one that meets the following conditions:

[0093] a. The due date is later than the current date, and the balance is greater than 0. Data without a balance field defaults to a balance greater than 0.

[0094] b. Filter out companies that are currently overdue, have poor credit records, or have abnormal business operations according to Dazhihui (a business information platform);

[0095] c. The company must be located in the city where the bank branch is located.

[0096]

[0097] (2) Business opportunity recommendation model construction

[0098] Define business opportunities and recommend points

[0099] in: The opportunity merging coefficient indicates whether the company has simultaneously captured other business opportunities (such as existing customers or high-quality customers);

[0100] Where: K: Number of business opportunity types; I k : Whether the k-th type of business opportunity has been hit; λ k : Weighting coefficients for various superiors; δ: Merging coefficient amplification factor.

[0101] Distance coefficient, representing the distance between the business and the bank branch;

[0102]

[0103] The company's own coefficient is composed of the following weighted sub-items:

[0104]

[0105] in: Enterprise type coefficient (e.g., state-owned enterprises, central enterprises, private enterprises, etc.); Enterprise size coefficient (e.g., large, medium, small, etc.); Enterprise qualification coefficient (such as listed companies, bond-issuing companies, and companies on high-quality lists); Industry coefficients (such as manufacturing, technology, finance, etc.); ω1+ω2+ω3+ω4=1: weights of each dimension.

[0106] (3) Business opportunity ranking mechanism

[0107] All business opportunities are sorted in descending order of referral score; if referral scores are the same, they are sorted in ascending order of distance.

[0108]

[0109] Step 3: Calculate and sort business opportunity recommendations.

[0110] Based on the financing maturity event opportunity model, a recommendation score for financing maturity event opportunities is determined, and a priority opportunity list is generated.

[0111] (1) Coefficient assignment and calculation

[0112] The model employs a combination of expert evaluation and historical data training to assign values ​​to each coefficient, and then dynamically optimizes the model using a machine learning approach.

[0113] For example, using a linear regression model to predict recommendation scores:

[0114]

[0115] Where θ = [θ0, θ1, θ2, θ3] are the model parameters, which are obtained through supervised learning training.

[0116] (2) Business opportunity screening mechanism

[0117] Set the recommendation score threshold R th Retain business opportunities with recommendation scores higher than this threshold:

[0118] Qualified = {f j |R j>R th}

[0119] (3) Output business opportunity list

[0120] The final output is a list of prioritized business opportunities after filtering. Where k≤n, sorted by recommendation score.

[0121] Step 4: Design of an Intelligent Business Opportunity Push Mechanism

[0122] Based on the business opportunity recommendation and distribution mechanism for financing maturity events, business opportunities are pushed to relevant banks through automatic or manual allocation methods.

[0123] (1) Push method selection

[0124] Automatic allocation: Target customers are associated with existing customers. Based on the customer relationship graph, they are matched and assigned to the organization of the existing customer with the highest correlation.

[0125] Manual allocation: Business opportunities are distributed layer by layer from the bank's head office, branches, and sub-branches.

[0126] (2) Customer Relationship Matching Algorithm

[0127] Construct a correlation matrix M between existing customers and target companies ij ,in:

[0128] Where: i: Existing customer ID; j: Target company ID; r ijk ω represents the degree of correlation of the k-th type of relationship (e.g., equity relationship, guarantee relationship, relationship within the same industrial park, etc.); k : Weight of each type of relationship; K: Total number of each type of relationship.

[0129] Select the business units of existing customers with the highest relevance to push business opportunities:

[0130]

[0131] (3) Push execution and feedback mechanism: After the push, record data such as the business opportunity marketing response and business opportunity conversion rate for model iteration and optimization.

[0132] Beneficial effects:

[0133] 1. Achieve integration and structured modeling of multi-source heterogeneous data to improve the comprehensiveness and accuracy of financing event identification.

[0134] Existing technologies largely rely on single-source data from internal bank systems (such as CRM and credit systems), failing to obtain customer financing information from external institutions. This invention introduces multi-source publicly available data (such as business registration information, judicial records, financial statements, public opinion dynamics, bond financing, trust financing, etc.), combined with natural language processing (NLP) and information extraction techniques, to construct a structured database of financing maturity events. This effectively identifies customer financing maturity events at other banks or non-bank financial institutions, overcoming the information blind spots of traditional methods and significantly improving the breadth and accuracy of financing event identification.

[0135] 2. Construct a business opportunity recommendation model that integrates multi-dimensional features to achieve dynamic evaluation and intelligent ranking of business opportunities.

[0136] Traditional methods lack comprehensive consideration of customer characteristics, resulting in a lack of targeted business opportunity identification. This invention innovatively constructs a "financing maturity event business opportunity recommendation model," which takes multiple dimensions such as enterprise attributes (e.g., enterprise nature, size, qualifications, industry), geographical location, and business opportunity merging coefficient as model inputs. Through a combination of expert scoring and machine learning, dynamic weighting and optimization are performed to generate business opportunity recommendation scores. Business opportunities are then ranked according to the recommendation scores to ensure that high-priority business opportunities are pushed first.

[0137] 3. Introduce customer relationship graphs and multi-dimensional matching algorithms to achieve intelligent push and precise targeting of business opportunities.

[0138] Existing systems suffer from "indiscriminate marketing and universal push" in the opportunity delivery process, leading to resource waste and inefficiency. This invention addresses this by constructing a customer relationship graph and utilizing multi-dimensional matching algorithms (such as equity relationships, guarantee relationships, and industrial park affiliations) to analyze the correlation between target companies and existing customers. This allows for precise matching of opportunities to the most suitable business unit (such as a branch or account manager), thereby achieving an intelligent push mechanism that prioritizes "who is most likely to follow up and who should be prioritized."

[0139] 4. Achieve full-process automation, reduce labor costs, and improve operational efficiency.

[0140] Traditional business opportunity discovery relies on manual due diligence and regular follow-ups by account managers, which is inefficient and difficult to scale. This invention constructs an automated processing workflow covering data collection, cleaning, modeling, business opportunity identification, recommendation ranking, and push execution, achieving end-to-end automated processing from data to business opportunities. This greatly reduces manual intervention and improves overall operational efficiency, making it particularly suitable for small and medium-sized banks or branches with limited resources.

[0141] 5. Supports multi-dimensional data fusion and intelligent decision-making, enhancing the commercial value of business opportunities and risk control capabilities.

[0142] This invention goes beyond simply identifying financing maturity events. It also supports the integration of multi-source heterogeneous data, such as corporate financial data, industry trends, and public opinion information, into a comprehensive analysis using AI models to assist banks in making intelligent decisions. For example, it can further assess a company's refinancing intentions, potential credit risks, and industry trends, thereby enhancing the commercial value of business opportunities and risk control capabilities.

[0143] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying and pushing business opportunities based on publicly available data regarding financing maturity events, characterized in that... The push method includes: Step S1: Collect and structure the heterogeneous data from multiple sources; Step S2: Construct a business opportunity model for financing maturity events; Step S3: Calculate and sort the business opportunity recommendations and output them; Step S4: Build an intelligent business opportunity push mechanism.

2. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 1, characterized in that, Step S1: Multivariate heterogeneous data acquisition and structured modeling specifically includes: Collect enterprise-related data from multiple publicly available data sources; Data extraction and cleaning: Natural Language Processing (NLP) and information extraction techniques are used to process unstructured text and extract key fields; Building a financing maturity event database: All extracted financing event data are structured and stored to build a financing maturity time database. Each record is represented as follows: Among them, E j For entities; For financing type; Due date; The amount of funding; The state is...

3. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 2, characterized in that, The collection of enterprise-related data from multiple publicly available data sources specifically includes: Business registration data: registered capital, legal representative information, shareholder structure; Judicial data: information on those subject to enforcement, records of dishonesty, and legal proceedings; Financial data: Annual report data, including balance sheet, income statement, and cash flow statement; Public opinion data: news, announcements, and social media information; Data on violations and penalties: tax violations, administrative penalties; Financing data: bank loans, trust financing, bond financing, financial leasing, and accounts receivable.

4. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 2, characterized in that, The key fields specifically include: financing amount, maturity date, and financing type.

5. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 1, characterized in that, Step S2: Constructing a financing maturity event business opportunity model specifically includes: Based on the database of financing maturity events Construct a business opportunity model for financing maturity events, including: identifying eligible financing maturity event business opportunities, and using the company's own coefficient, distance coefficient, and business opportunity merging coefficient as model coefficients to form a business opportunity recommendation score.

6. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 5, characterized in that, The database based on financing maturity events The specific steps in constructing a business opportunity model for financing maturity events include: Business opportunity identification criteria: For database Each record f in j A business opportunity is considered to be one that meets the following conditions: The due date is later than the current date and the balance is greater than 0. Data without a balance field defaults to a balance greater than 0. Filter out companies that are currently overdue, have poor credit records, or have abnormal business operations according to Dazhihui (a business information platform). Businesses must be located in the city where the bank's branch is located; Business Opportunity Recommendation Score Model Construction: Defining Business Opportunity Recommendation Score in: The opportunity merging coefficient indicates whether the company has simultaneously captured other business opportunities (such as existing customers or high-quality customers); Where K represents the number of business opportunity types; I represents the number of business opportunity types. k : Whether the k-th type of business opportunity has been hit; λ k δ: Weighting coefficients for various superior levels; δ: Merging coefficient amplification factor; Distance coefficient, representing the distance between the business and the bank branch; The company's own coefficient is composed of the following weighted sub-items: in: Enterprise nature coefficient; Enterprise size coefficient; Enterprise qualification coefficient; Industry coefficient; ω1+ω2+ω3+ω4=1: weights of each dimension; Business opportunity ranking mechanism: All business opportunities are sorted in descending order of referral score; if the referral scores are the same, they are sorted in ascending order of distance.

7. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 1, characterized in that, Step S3: Calculation and ranking of business opportunity recommendations specifically includes: Based on the financing maturity event opportunity model, a recommendation score for financing maturity event opportunities is determined, and a priority list of opportunities is generated, including: Coefficient assignment and calculation: A combination of expert evaluation and historical data training is used to assign values ​​to each coefficient in the model, and dynamic optimization is performed through machine learning models. Business opportunity screening mechanism: setting a recommendation score threshold R th Retain business opportunities with a referral score higher than this threshold: Qualified = {f j |R j >R th }; Output Business Opportunity List: Outputs a filtered list of priority business opportunities. Where k≤n, sorted by recommendation score.

8. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 1, characterized in that, Step S4: Constructing an intelligent business opportunity push mechanism specifically includes: Based on the business opportunity recommendation and distribution mechanism for financing maturity events, business opportunities are pushed to relevant banks through automatic or manual allocation methods.

9. The method for identifying and pushing business opportunities based on publicly available data of financing maturity events according to claim 8, characterized in that, The mechanism for recommending and pushing business opportunities based on financing maturity events, and for reaching out to potential clients, specifically includes: Push method selection includes: Automatic allocation: Target customers are associated with existing customers. Based on the customer relationship graph, they are matched and assigned to the organization of the existing customer with the highest correlation. Manual allocation: Business opportunities are distributed layer by layer from the bank's head office, branches, and sub-branches; The customer relationship matching algorithm includes: constructing a correlation matrix M between existing customers and target companies. ij ,in: Where: i is the existing customer ID; j is the target company ID; r ijk ω represents the degree of association of the k-th type of relation; k represents the weight of each type of relation; K represents the total number of relation types. Select the business units of existing customers with the highest relevance to push business opportunities: The push execution and feedback mechanism includes: recording the marketing response status and conversion rate data after the push is sent, which is used for model iteration and optimization.

Images