Enterprise operation state scoring model construction method and device, equipment and medium

By extracting features from the multi-source enterprise information integration database and building an enterprise operating status scoring model, the problem of high dependence on industrial and commercial information and labor costs in the existing technology is solved, and a high accuracy and efficient enterprise operating status evaluation is achieved.

CN118941158APending Publication Date: 2024-11-12CHINA MOBILE ONLINE SERVICES CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411126958.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing methods for identifying business status are heavily dependent on the quality of industrial and commercial information and a large amount of labor costs, resulting in inaccurate assessment, low investigation efficiency and insufficient resource utilization.

Method used

Select industrial and commercial features, Internet features, hotline traffic features and integration quality features from the multi-source enterprise information integration database, build a feature engineering sample, and build an enterprise operating status scoring model based on these features to automatically score the enterprise's operating status.

Benefits of technology

It achieves a high-accurate business status score without manual participation, improves data processing efficiency and resource utilization, and ensures the accuracy and efficiency of evaluation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941158A_ABST
    Figure CN118941158A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of artificial intelligence, and provides a construction method and device of an enterprise operation state scoring model, equipment and a medium. The method comprises the following steps: selecting industrial and commercial characteristics, internet characteristics, hotline traffic characteristics and integration quality characteristics of an enterprise from an enterprise information integration library corresponding to multi-source enterprise information; constructing a feature engineering sample according to the industrial and commercial features, the Internet features, the hotline traffic features and the integrated quality features; and according to the feature engineering sample, an enterprise operation state scoring model is constructed, and the enterprise operation state scoring model is used for scoring the operation state of the to-be-verified target enterprise. According to the method, the enterprise operation state scoring model with relatively high accuracy can be constructed based on the enterprise information integration library corresponding to the multi-source enterprise information without manual participation, so that the operation state score of the to-be-verified target enterprise can be automatically and accurately determined, and the resource utilization rate is improved while the data processing efficiency is improved in the whole process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and medium for constructing a business operation status scoring model. Background Art

[0002] With the advent of the digital age, the authenticity of a company's operating status is crucial to the decision-making and marketing of frontline government and enterprise businesses. However, due to the diversity of corporate information sources and the lack of timely updates, it is difficult to accurately identify the authenticity of a company's operating status before conducting government and enterprise marketing.

[0003] Existing methods for identifying the business status of enterprises often use industrial and commercial information and manual full-scale surveys to obtain the true business status of enterprises. Specifically, the business information of enterprises is obtained, and the business status of enterprises is used to determine whether the enterprises are in operation. The data of enterprises in operation are then manually surveyed, and then the enterprises that are manually identified as having normal business status are applied to government-enterprise marketing to assist offline contacts and online cross-marketing. The entire process is heavily dependent on the quality of industrial and commercial information itself and a large amount of manpower costs, which can easily lead to inaccurate business status of enterprises in the final assessment, and problems such as large survey volume, low efficiency and insufficient resource utilization. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment and medium for constructing an enterprise business status scoring model, which is used to solve the technical problems that the existing enterprise business status identification method is heavily dependent on the quality of the industrial and commercial information itself and a large amount of manpower costs, which can easily lead to the final evaluation of the enterprise business status is not accurate enough, and is also prone to problems such as large amount of investigation, low efficiency and insufficient resource utilization.

[0005] In a first aspect, an embodiment of the present application provides a method for constructing a business operation status scoring model, comprising the following steps.

[0006] From the enterprise information integration database corresponding to multi-source enterprise information, select the enterprise's industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integration quality characteristics; Constructing a feature engineering sample according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics, and the integrated quality characteristics; A business operation status scoring model is constructed based on the feature engineering samples, and the business operation status scoring model is used to score the business status of the target enterprise to be verified.

[0007] In one embodiment, constructing an enterprise business status scoring model based on the feature engineering sample includes: obtaining a preset enterprise, wherein the preset enterprise contains labeled valid enterprise features; constructing multiple base decision trees based on the feature engineering sample and the labeled valid enterprise features; and constructing the enterprise business status scoring model based on the multiple base decision trees.

[0008] In one embodiment, the method further includes: extracting structural features of multi-source enterprise information to obtain multi-source enterprise structural information; determining enterprise features matching preset enterprises based on the multi-source enterprise structural information; and constructing the enterprise information integration library based on the enterprise features.

[0009] In one embodiment, the preset enterprise includes at least one preset sub-enterprise, and determining the enterprise characteristics matching the preset enterprise based on the multi-source enterprise structure information includes: performing the following operations for any preset sub-enterprise of the at least one preset sub-enterprise: calculating the first Internet characteristics and the first hotline traffic characteristics of each enterprise structure information in the multi-source enterprise structure information; determining the second Internet characteristic with the highest similarity to the preset sub-enterprise from all the first Internet characteristics, and determining the second hotline traffic characteristic with the highest similarity to the preset sub-enterprise from all the first hotline traffic characteristics; determining the integrated quality characteristics of the preset sub-enterprise based on the second Internet characteristic and the second hotline traffic characteristic; and determining the second Internet characteristic, the second hotline traffic characteristic and the integrated quality characteristic as the enterprise characteristics of the preset sub-enterprise.

[0010] In one embodiment, constructing multiple base decision trees based on the feature engineering samples and the labeled effective enterprise features includes: constructing an original feature set based on the feature engineering samples and the labeled effective enterprise features; obtaining multiple training subsets from the original feature set by random sampling with replacement, the number of features in each training subset being the same as the number of features in the original feature set; and constructing corresponding base decision trees based on the multiple training subsets.

[0011] In one embodiment, the calculation of the first Internet feature and the first hotline traffic feature of each enterprise structure information in the multi-source enterprise structure information includes: performing the following operations for each enterprise structure information: determining the first Internet feature based on the industrial and commercial enterprise name, enterprise identity, contact number, detailed address and geographic coordinate information in the enterprise structure information; determining the first hotline traffic feature based on the industrial and commercial enterprise name, the enterprise identity and the contact number.

[0012] In a second aspect, an embodiment of the present application provides a device for constructing a business operation status scoring model, comprising the following modules.

[0013] A feature determination module is used to select the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information; and construct a feature engineering sample based on the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics; The model building module is used to build an enterprise business status scoring model based on the feature engineering sample, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a transceiver, and a processor. The memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations.

[0015] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, the steps of the method for constructing a business operation status scoring model described in the first aspect are implemented.

[0016] In a fifth aspect, an embodiment of the present application further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for constructing a business operation status scoring model described in the first aspect.

[0017] In a sixth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for constructing the enterprise business status scoring model described in the first aspect.

[0018] The method, device, equipment and medium for constructing the enterprise business status scoring model provided in the embodiment of the present application select the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information; construct feature engineering samples according to the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics; construct an enterprise business status scoring model according to the feature engineering samples, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified. This method does not require manual participation, and can construct a high-accuracy enterprise business status scoring model based on the enterprise information integration library corresponding to the multi-source enterprise information, and can then automatically and accurately determine the business status score of the target enterprise to be verified. The whole process improves the data processing efficiency while improving resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 It is a flow chart of a method for identifying business operation status provided by the prior art.

[0021] Figure 2 It is a flowchart of a method for constructing a business operation status scoring model provided in an embodiment of the present application.

[0022] Figure 3 It is a structural diagram of a device for constructing an enterprise business status scoring model provided in an embodiment of the present application.

[0023] Figure 4 This is one of the structural schematic diagrams of the electronic device provided in the embodiment of the present application.

[0024] Figure 5 This is the second structural schematic diagram of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0026] In order to better understand the embodiments of the present application, the prior art is first described in detail.

[0027] Figure 1 This is a flow chart of the business status identification method provided by the prior art. Figure 1 The method for identifying the business status of an enterprise includes: step 101, obtaining the business information of the enterprise, and judging whether the enterprise is in operation based on the business status; step 102, delegating the data of the operating enterprises to the account manager and the government and enterprise special seat for manual investigation; step 103, applying the enterprises manually identified as having normal business status to government and enterprise marketing, to assist offline contact and online cross-marketing.

[0028] There are mainly three problems with the above-mentioned method of identifying the business status of an enterprise.

[0029] Question 1: Using the business status of an enterprise based on the industrial and commercial registration information to identify the authenticity of the enterprise's business status relies on the enterprise's own declaration and the review of the registration authority, ignoring problems such as information asymmetry caused by untimely updates of industrial and commercial information. It does not have the real-time advantage of Internet information and can easily result in a low accuracy rate in identifying the enterprise's business status.

[0030] Question 2: The relationship between other multi-dimensional information such as enterprise development information (such as the company's expansion plan, new product research and development, and technological innovation), risk information (such as financial risk, market risk, and legal risk), and activity information (such as change information, equity information, and hotline call information) and the company's actual operating status has not been explored. There is a possibility of information loss, and it is impossible to provide a more comprehensive and accurate assessment of the company's operating status.

[0031] Question 3: The operating enterprises identified through the experience status of industrial and commercial enterprises cannot be directly used in government-enterprise marketing. They need to be manually verified again, and the survey can only identify whether the enterprise exists or not (0, 1). The degree of quantification of the enterprise operation status assessment results is not specific enough and has poor interpretability, which not only causes a serious waste of resources, but also has the problem of low survey efficiency and poor usability of survey results.

[0032] To sum up, the above-mentioned methods for identifying the business status of enterprises rely heavily on the quality of industrial and commercial information itself and a large amount of manpower costs, which may easily lead to inaccurate final assessment of the business status of enterprises, and may also easily lead to problems such as large amount of investigation, low efficiency and insufficient resource utilization.

[0033] In order to solve the above technical problems, the method for constructing an enterprise business status scoring model provided in the embodiment of the present application selects the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information; constructs a feature engineering sample based on the business characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics; constructs an enterprise business status scoring model based on the feature engineering sample, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified. This method does not require manual participation, and can construct a high-accuracy enterprise business status scoring model based on the enterprise information integration library corresponding to the multi-source enterprise information, and can then automatically and accurately determine the business status score of the target enterprise to be verified. The entire process improves data processing efficiency while improving resource utilization.

[0034] The following describes in detail the method for constructing the enterprise business status scoring model provided in the embodiment of the present application by taking electronic equipment as an example.

[0035] Figure 2 Schematic diagram of the process of constructing the enterprise business status scoring model provided in the embodiment of the present application. Figure 2, an embodiment of the present application provides a method for constructing an enterprise business status scoring model, including the following steps 201-203.

[0036] Step 201: Select the business characteristics, Internet characteristics, hotline traffic characteristics and integration quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information.

[0037] Among them, multi-source enterprise information refers to various enterprise information about the same enterprise or different enterprises from different channels, different platforms or different systems.

[0038] An enterprise information integration library refers to a data warehouse specially designed to store, manage, integrate and analyze enterprise information from multiple different sources.

[0039] Industrial and commercial characteristics refer to the production and operation-related characteristics of an enterprise as an economic entity, which can be expressed as f gs express.

[0040] Internet characteristics refer to the relevant characteristics of enterprises registered on computer networks, which can be net express.

[0041] Hotline traffic characteristics refer to the traffic-related characteristics of enterprise communications, which can be call express.

[0042] The integrated quality feature is a fusion feature of Internet features and hotline traffic features, which can be used s express.

[0043] Optional, business characteristics gs Can include: registered capitalf regm 、Paid-in capital paid , Staff size pers and the number of insured persons insured 4 continuous features, and the establishment time f time 、Enterprise Type type , Taxpayer qualifications taxpayer 、Branches branch and annual reports report There are five discrete features, including continuous features, which are directly expressed by integer, floating point and other numerical values; discrete features are expressed by analyzing their derived features and numerical values.

[0044] Based on this, the business characteristics can be used gs ={f regm ,f paid ,f pers ,f insured ,f time ,f type ,f taxpayer ,fbranch ,f report}express.

[0045] Optional, Internet feature f net Can include: Tax class f tax 、Change Record chg 、Outward Investment inv 、Whether to be listed list 、Court Notice court 、Dishonest debtor debtor 、Equity pledge pledge 、Movable property mortgage mortg 、Administrative penalties pena 、Tendering bid 、Recruitment information recru 、Financing information finance 、Intellectual property information intel 、Administrative license asmin and qualification certificates cert And so on 15 discrete features.

[0046] Based on this, Internet features can be used net ={f tax ,f chg ,f inv ,f list ,f court ,f debtor ,f pledge ,f mortg ,f pena ,

[0047] f bid ,f recru ,f finance ,f intel ,f asmin ,f cert}express.

[0048] Optional, hotline feature call Can include: Number Star rating f star , Arrears information fee and complaint information co And other three discrete features.

[0049] Based on this, the hotline traffic characteristics can be used net ={f star ,f fee ,f co}express.

[0050] The discrete characteristics of the above-mentioned industrial and commercial characteristics, Internet characteristics and hotline traffic characteristics are explained in detail.

[0051] (1) Discrete characteristics in industrial and commercial characteristics

[0052] Establishment time: expressed as a time variable, that is, the difference from the current date.

[0053] In view of multi-classification features, the enterprise type and taxpayer qualification are digitized in the form of label coding. Among them, the enterprise type is divided into joint venture, sole proprietorship, state-owned, private, state-owned, collectively owned, joint-stock and limited liability, etc., and the taxpayer qualification is divided into general VAT taxpayers and small-scale VAT taxpayers and other multi-category features. A large number of mapping methods are used to assign a unique digital code to each classification, such as: "joint venture, sole proprietorship, state-owned, private, state-owned, collectively owned, joint-stock and limited liability" can be represented by [0,1,2,3,4,5,6,7] after encoding.

[0054] Branches are represented by the number of branches. For example, if the enterprise has 3 branches, the characteristic value is 3, and if there are no branches, the characteristic value is 0.

[0055] Corporate annual report: It is expressed by the time difference between the latest annual report release and the latest annual report release, such as current year - latest annual report year = 2024-2023 = 1 (year), that is, the latest annual report is nearly one year.

[0056] (3) Discrete characteristics in Internet characteristics

[0057] Tax grade: Taxpayer qualifications are divided into five levels: A, B, M, C and D, corresponding to the annual evaluation index scores [90, 100], [70, 90], (70, 100], [40, 70) and [0, 40]. It is quantified by the minimum boundary value, that is, the A-level range is quantified as 90, the B-level range is quantified as 70, and the M-level range overlaps with the B-level range and the A-level range. Since the M-level range is in line with the conditions of newly established enterprises and is re-rated in the target month of the next year (such as March-April), if the credit status of the enterprise is good, it is possible to upgrade to the B-level range. Therefore, the M-level range is quantified as 65, slightly lower than the B-level range, the C-level range is quantified as 40, and the D-level range is quantified as 0.

[0058] Change records are expressed by the number of changes in the past year.

[0059] Outward investment can be the number of outward-invested enterprises that are operating normally, or the total subscribed capital of outward-invested enterprises.

[0060] Court announcements can be the number of court announcements received or the number of judicial cases.

[0061] The dishonest debtor can be the number of times the main personnel and legal person have been executed for dishonesty, or the time difference between the most recent dishonest execution and the current time.

[0062] The pledge of equity can be the total number of pledged equity in the past year, or the time difference between the most recent pledge of equity and the current time.

[0063] Movable property mortgage can be the number of movable property mortgages in the past year or the total amount of movable property mortgages.

[0064] Administrative penalties can be the number of administrative penalties in the past year.

[0065] Bidding and tendering can be the number of bidding and winning bids in the past two years, the bidding and winning bid amounts, or the time difference between the bidding or winning bid and the current time.

[0066] Recruitment information can be the number of recruitment positions or the number of people to be recruited.

[0067] Intellectual property information can be the total amount of intellectual property rights such as software copyrights and patents obtained, or the number of intellectual property rights obtained in the past year.

[0068] Qualification certificates, represented by the number of qualification certificates.

[0069] Administrative license can be whether the administrative license is renewed or the number of administrative licenses; listing information is represented in binary form based on whether the company is listed.

[0070] Financing information, feature the financing rounds by equidistant binning. For example, define the following bins: early rounds, mid-term rounds, and late rounds. Among them, early rounds include seed rounds, angel rounds, etc.; mid-term rounds include rounds A, A+, B, etc.; late rounds include rounds C, D, and later rounds. Then, assign a unique integer value to each bin as the feature value. In this way, the financing round, which was originally an ordered sequence feature, is converted into a discrete feature.

[0071] (3) Discrete features in hotline traffic characteristics

[0072] The number star rating is represented by a numerical value from 1 to 5.

[0073] The arrears information can be the arrears amount in the past year or the number of times the payment was made.

[0074] Complaint information can be the number of complaints in the past year or the complaint satisfaction.

[0075] It should be noted that, since there is a large amount of enterprise information in the enterprise information integration library, the industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics finally obtained by the electronic device are also large. In this way, step 201 introduces multiple information sources, so that the enterprise information integration library corresponding to the multi-source enterprise information is diversified, which can provide enterprises with more comprehensive and detailed business status information, so that the business status of the enterprise can be judged more accurately in the future.

[0076] In some embodiments, before step 201, the method may further include: the electronic device extracts structural features of multi-source enterprise information to obtain multi-source enterprise structural information; the electronic device determines enterprise features that match a preset enterprise based on the multi-source enterprise structural information; the electronic device constructs an enterprise information integration library based on the enterprise features.

[0077] Among them, multi-source enterprise information can also be called multi-source enterprise heterogeneous information.

[0078] In the process of building an enterprise information integration library, electronic devices can determine multi-source enterprise information (such as enterprise industrial and commercial information, Internet information, and government-enterprise hotline call information, etc.) through different channels and strategies; then, natural language processing (NLP) technology is used to extract enterprise information / structural features from multi-source enterprise information, and the enterprise name or enterprise identity (Identification, ID) of the same channel is used as a unique identifier for filtering and deduplication, so as to convert the multi-source enterprise information into multi-source enterprise structural information (such as multi-source enterprise names (industrial and commercial registration names, storefront names and former names, etc.), enterprise IDs (such as unified social credit codes, organization codes, industrial and commercial registration numbers and taxpayer identification numbers, etc.), contact numbers, detailed addresses (such as industrial and commercial registration addresses, Internet identification addresses, etc.) and geographic coordinate information, etc.).

[0079] Next, feature calculations are performed on the multi-source enterprise structure information to determine enterprise features that match the preset enterprise. It should be noted that since the preset enterprise includes at least one preset sub-enterprise, the number of enterprise features obtained is also at least one, and then a more comprehensive enterprise information integration library is constructed based on all enterprise features.

[0080] The above-mentioned entire integrated library construction process focuses on three channels: enterprise industry and commerce, the Internet, and hotlines. By integrating Internet information with high timeliness, government-enterprise hotline traffic information, and enterprise industry and commerce information, the enterprise information integrated library is constructed, which can expand and enhance the multi-dimensional attribute information of enterprises, and obtain attribute values ​​that measure the quality of multi-source enterprise information integration, avoid errors and losses of enterprise information, and at the same time, make up for the problem of poor timeliness of industrial and commercial information, and solve the quality problem of industrial and commercial information to a certain extent. In addition, compared with the existing technology for the identification of industrial and commercial registration status, the above-mentioned enterprise information integrated library not only considers traditional financial characteristics such as registered capital and paid-in capital, but also introduces more abundant characteristics such as staff size, number of insured persons, enterprise type, enterprise annual report, longitude and latitude, Internet status, hotline traffic activity, and information integration quality. These characteristics can more comprehensively reflect the operating status, scale, market activity, and integrated information quality of enterprises, and provide more basis for judging whether enterprises are in operation. At the same time, compared with the lag of business registration information in the existing technology, the above-mentioned entire integrated library construction process can obtain and process various information from channels such as the Internet and hotlines in real time, thereby realizing dynamic monitoring of the business status of enterprises. This means that the actual business situation of enterprises can be reflected more timely, and more valuable decision-making support can be provided to enterprises and relevant departments in the future.

[0081] Optionally, electronic devices can use targeted means to determine multi-source enterprise information through different channels and strategies. Targeted means may include: third-party procurement, Internet crawlers, data extraction-transformation-loading (ETL) real-time scheduling, and application programming interface (API) calls.

[0082] Optionally, the formats of the multi-source enterprise information may include text, pictures, audio and video, etc.

[0083] In some embodiments, a preset enterprise includes at least one preset sub-enterprise, and the electronic device determines enterprise characteristics that match the preset enterprise based on multi-source enterprise structure information, which may include: performing the following operations for any preset sub-enterprise in at least one preset sub-enterprise: the electronic device calculates the first Internet characteristics and the first hotline traffic characteristics of each enterprise structure information in the multi-source enterprise structure information; the electronic device determines the second Internet characteristic with the highest similarity to the preset sub-enterprise from all the first Internet characteristics, and determines the second hotline traffic characteristic with the highest similarity to the preset sub-enterprise from all the first hotline traffic characteristics; the electronic device determines the integrated quality characteristics of the preset sub-enterprise based on the second Internet characteristic and the second hotline traffic characteristic; the electronic device determines the second Internet characteristic, the second hotline traffic characteristic and the integrated quality characteristic as the enterprise characteristics of the preset sub-enterprise.

[0084] Among them, the preset enterprise refers to the enterprise that has been registered with the industrial and commercial administration.

[0085] In the process of determining the enterprise characteristics of each preset sub-enterprise, the first Internet characteristics and the first hotline traffic characteristics of each enterprise structure information in the multi-source enterprise structure information can be calculated first, so that multiple first Internet characteristics and multiple first hotline traffic characteristics are obtained; then the information is integrated through the similarity score, specifically, the similarity between all the first Internet characteristics and the Internet characteristics of the preset sub-enterprise is determined, and the second Internet characteristics with the highest similarity are determined from all the first Internet characteristics. Similarly, the similarity between all the hotline traffic characteristics and the hotline traffic characteristics of the preset sub-enterprise is determined, and the second hotline traffic characteristics with the highest similarity are determined from all the first hotline traffic characteristics. At this time, the electronic device can calculate the second Internet characteristics and the second hotline traffic characteristics to obtain the integrated quality characteristics of the preset sub-enterprise, and then determine the second Internet characteristics, the second hotline traffic characteristics and the integrated quality characteristics as the enterprise characteristics of the preset sub-enterprise. Based on this, the enterprise characteristics of each preset sub-enterprise can be determined.

[0086] It should be noted that in the process of calculating the similarity, the similarity between the geographic coordinate information is calculated by 1 / (1+Euclidean distance), and other features such as company name, company ID, contact number, and detailed address are calculated using the cosine similarity algorithm.

[0087] The Euclidean distance refers to the distance between the geographic coordinate information in the enterprise structure information and the geographic coordinate information of the preset subsidiary enterprise.

[0088] In some embodiments, the electronic device calculates the first Internet feature and the first hotline traffic feature of each enterprise structure information in the multi-source enterprise structure information, which may include: performing the following operations for each enterprise structure information: the electronic device determines the first Internet feature based on the industrial and commercial enterprise name, enterprise identity, contact number, detailed address and geographic coordinate information in the enterprise structure information; the electronic device determines the first hotline traffic feature based on the industrial and commercial enterprise name, enterprise identity and contact number.

[0089] In the process of determining the first Internet feature of each enterprise structure information, the first Internet feature can be determined based on the business name, enterprise identity, contact number, detailed address and geographic coordinate information in the enterprise structure information, and the weights corresponding to these five data. Specifically, the first formula can be used to determine the first Internet feature. Calculated, Indicates the first Internet feature, Represents the characteristics of the i-th data among the above five data, Represents the weight corresponding to the feature of the i-th data.

[0090] In the process of determining the first hotline call characteristics of each enterprise structure information, the business name, enterprise identity and contact number in the enterprise structure information and the three data

[0091] The corresponding weights are used to determine the first hotline traffic characteristics. Specifically, the second formula can be used Calculated, Indicates the characteristics of the first hotline traffic. Represents the characteristics of the jth data among the above three data, Represents the weight corresponding to the feature of the j-th data.

[0092] Optionally, the above integrated quality characteristics can be represented by the third formula S=a× +b× It is calculated that S represents the integrated quality characteristic, a represents the first constant, and b represents the second constant. Usually, a=b=1 / 2.

[0093] Optionally, the electronic device constructs an enterprise information integration library according to enterprise characteristics, which may include: the electronic device constructs an initial enterprise information integration library according to enterprise characteristics; the electronic device uses a general anomaly detection model to eliminate abnormal data from the initial enterprise information integration library to obtain the enterprise information integration library.

[0094] Among them, the general anomaly detection model can be used ex Indicates that it is used to identify abnormal enterprises.

[0095] Optionally, the general anomaly detection model may include: a single feature anomaly detection model M s and multi-feature anomaly detection model M m Based on this, the single feature anomaly detection model M s , can quickly discover and deal with obvious abnormal enterprises; through the multi-feature anomaly detection model M m , which can more deeply explore abnormal enterprises hidden in the combined information of multiple enterprises.

[0096] Optionally, build a single feature anomaly detection model M based on business needs analysis s , the single feature anomaly detection model M s May include: Business status R sts For revocation, cancellation, or relocation of the business name R name The length is less than 5 characters or the same as the legal person; Industrial and commercial enterprise identity R no Empty; Industrial and Commercial Administrative District R addr Wrongly assigned; registered address R distIncludes mobile operations, cluster registration, managed registration, and business secretary hosting.

[0097] Based on this, a single feature anomaly detection model can be used

[0098] M s =R sts R name R no R addr R dis R long R lat express.

[0099] Among them, R long Represents latitude characteristics; R lat Represents the longitude feature.

[0100] Optionally, build a multi-feature anomaly detection model M based on business needs analysis m , the multi-feature anomaly detection model M m May include: Business registration status R gssts The Internet enterprise is still in existence (in operation, opened for business, registered) but the status is R netsts For suspension of business, closure of business; the registered address of the enterprise is R gsaddr Not empty but Internet latitude and longitude (actual business address) R netaddr Empty; there is a corporate legal person R gstel But the contact number R calltel The number is empty, suspended or has no consumption record.

[0101] Based on this, the multi-feature anomaly detection model can be used

[0102] M m =(R gssts R netsts ) (R gsaddr R netaddr ) (R gstel R calltel )express.

[0103] The entire process of building the above-mentioned general anomaly detection model can effectively eliminate invalid enterprises, avoid the interference of a large number of invalid enterprises on the business status prediction scoring algorithm, and to a certain extent solve the problems of key feature recognition errors and low prediction result accuracy caused by a large number of invalid enterprises.

[0104] Step 202: construct a feature engineering sample based on industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics.

[0105] Among them, the feature engineering samples can be used Mix =f gs f net f call f s express, Represents a union.

[0106] Since there are a large number of industrial and commercial features, Internet features, hotline traffic features, and integrated quality features, the feature engineering samples finally constructed also contain a large number of sub-feature engineering samples f mix , at this time, each sub-feature engineering sample f mix ={f1,f2,…,f Y}, f Y Represents the Yth enterprise characteristic, Y≥1.

[0107] In order to ensure that the enterprise business status scoring model can correctly interpret the input feature values ​​during training, each sub-feature engineering sample f mix All enterprise characteristics in are normalized, specifically using Z-score standardization, namely: , .

[0108] in, represents the αth enterprise characteristic among Y enterprise characteristics; represents the standardized αth enterprise characteristic; represents the mean of Y enterprise characteristics; Represents the standard deviation of Y enterprise characteristics.

[0109] Step 203: construct an enterprise business status scoring model based on the feature engineering sample, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0110] In some embodiments, the electronic device constructs a business operation status scoring model based on feature engineering samples, which may include: the electronic device obtains a preset enterprise, and the preset enterprise contains labeled valid enterprise features; the electronic device constructs multiple base decision trees based on the feature engineering samples and the labeled valid enterprise features; the electronic device constructs the business operation status scoring model based on the multiple base decision trees.

[0111] In the process of constructing an enterprise business status scoring model, the electronic device can first obtain the labeled effective enterprise features; then, based on the feature engineering samples and the labeled effective enterprise features, construct multiple base decision trees, and then construct the enterprise business status scoring model, which is a random forest model.

[0112] It should be noted that existing technologies often judge whether an enterprise is in operation based on fixed standards or thresholds, and lack personalized consideration for different enterprise types and business models. The enterprise operating status scoring model constructed by the above process can learn and identify the differences between different enterprises, thereby realizing the determination of personalized operating status of different enterprises. This helps to more accurately reflect the actual operating conditions of enterprises and meet the specific needs of different enterprises. At the same time, the entire model building process has strong scalability, and more information sources and enterprise features can be added as needed to adapt to the ever-changing market environment and business needs. The random forest algorithm also has good flexibility and can adapt to enterprise information of different sizes and types, providing support for a wider range of application scenarios.

[0113] In addition, the above-mentioned enterprise operating status scoring model provides a method for evaluating the importance of features, which plays a key role in understanding which features are critical to determining whether an enterprise is operating. This transparency and explainability helps to enhance the reliability of the model, while providing deeper insights and explanations for enterprises and related departments, allowing them to better understand the operating status of the enterprise.

[0114] In some embodiments, the electronic device constructs multiple base decision trees based on feature engineering samples and labeled effective enterprise features, which may include: the electronic device constructs an original feature set based on the feature engineering samples and the labeled effective enterprise features; the electronic device obtains multiple training subsets from the original feature set by random sampling with replacement, and the number of features in each training subset is the same as the number of features in the original feature set; the electronic device constructs corresponding base decision trees based on the multiple training subsets.

[0115] In the process of building a business operation status scoring model, electronic devices can use bootstrap sampling to determine multiple training subsets. Specifically, the original feature set is constructed based on feature engineering samples and labeled valid enterprise features; then, based on the Y enterprise features in the original feature set, n training subsets are obtained by random sampling with replacement, where n ≥ 2, and each training subset is the same size as the original data set, but may contain repeated samples, among which the unselected enterprise features constitute the test set. In addition, the number of enterprise features in each training subset is y.

[0116] Next, for each training subset, the y enterprise features included in the training subset can form a feature subset, and the best split point is selected for splitting according to the principle of minimum node impurity. The split result is divided into sub-nodes, and the y enterprise features are divided into corresponding sub-nodes at the same time. The division is repeated layer by layer, and finally all the nodes of the decision tree are obtained to construct the base decision tree corresponding to the training subset.

[0117] All training subsets are traversed until n base decision trees are generated, and these n base decision trees are combined to form a random forest model, that is, a corporate operating status scoring model. In other words, for each training subset and the corresponding feature subset, a base decision tree can be constructed.

[0118] Optionally, after step 203, the method may further include: the electronic device calculates the global importance corresponding to each of the different enterprise features in the original feature set according to the enterprise business status scoring model; the electronic device determines the target enterprise features corresponding to the target global importance according to all global importances; the electronic device constructs features for the target enterprise to be verified according to the target enterprise features to obtain feature engineering to be verified; the electronic device inputs the feature engineering to be verified into the enterprise business status scoring model to obtain the business status score output by the enterprise business status scoring model.

[0119] After determining the enterprise business status scoring model, the electronic device can calculate the global importance of different enterprise features in the original feature set based on the enterprise business status scoring model. Each global importance is used to measure the contribution of the corresponding enterprise feature to the model prediction; then, the electronic device sorts all global importances, determines k global importances from large to small as the target global importance, k>1, and determines the target enterprise features corresponding to the target global importance from all enterprise features. The target enterprise features can construct a key feature subset; finally, after obtaining the target enterprise to be verified, the feature construction of the target enterprise to be verified can be carried out according to the key feature subset to obtain the feature engineering to be verified, and then input into the enterprise business status scoring model to obtain the business status score output by the enterprise business status scoring model. The whole process does not require human participation, and a high-accuracy business status score can be obtained, which solves the problem of information loss caused by relying on a single attribute, and the real business status of the enterprise is explicitly represented by a quantitative value, making the representation of the business status of the enterprise more explicit.

[0120] Optionally, the operating status score can be calculated using Score=P(β), where Score represents the operating status score; P represents the enterprise operating status scoring model; and β represents the feature engineering to be verified.

[0121] Specifically, the process of the electronic device calculating the global importance of each of the different enterprise features in the original feature set is as follows: the 10-fold cross-validation out-of-bag (OOB) error method is used as the performance evaluation of the random forest pre-training model to find the best parameter combination. Finally, the best parameter combination is selected as the optimal enterprise business status scoring model, and the global importance of each enterprise feature is calculated based on the enterprise business status scoring model. The specific steps are as follows: S1. Calculate the initial error: For each base decision tree, use the out-of-bag data to calculate the first initial error of the base decision tree (such as classification error rate or mean square error). S2. Shuffle the feature value: Randomly shuffle the feature value X of a certain enterprise in the out-of-bag data. c , keeping other enterprise characteristics unchanged, and then recalculating the error to obtain the second initial error. S3. Calculating the error difference: Calculate the difference between the first initial error and the second initial error. This difference reflects the enterprise characteristic X c The impact on the performance of the enterprise business status scoring model. S4. Cumulative error difference: For all base decision trees, the cumulative enterprise feature X c The corresponding error difference is calculated and the accumulated result is determined as the enterprise feature X c Based on this, the corresponding global importance of different enterprise features in the original feature set can be calculated.

[0122] That is to say, the entire model construction process described above inputs the annotated effective enterprise features (i.e., the effective enterprise features that have been surveyed and labeled) into the enterprise business status scoring model based on the feature engineering samples, obtains different feature subsets by using the self-service sampling method for the original feature set, randomly selects node attributes for each feature subset, and selects the optimal node attributes from them for splitting to form a base decision tree, and uses the majority result method to comprehensively analyze multiple unrelated base decision trees to obtain the final result, and uses the grid search method to iteratively optimize the four parameters of the number of all base decision trees, the maximum number of features, the maximum depth of the tree, and the minimum sample split to build a high-accuracy enterprise business status scoring model. In this way, when constructing and verifying the enterprise business status scoring model, cross-validation and grid search are used to repeatedly evaluate the performance of the model and select the optimal model parameters. By continuously tuning the model, the accuracy and stability of the judgment can be further improved to ensure the effectiveness and reliability of the enterprise business status scoring model in practical applications.

[0123] In summary, the effective enterprise features that have been identified through information integration and anomaly detection and have been surveyed and labeled are used as training sets to input into the enterprise business status scoring model for learning, and the importance weights of different features in the enterprise information are obtained; then, the recursive feature elimination method is used to extract the key feature subset corresponding to the enterprise business status scoring model; then, the optimal pre-training model is obtained through parameter optimization to obtain the enterprise business status scoring model; finally, the trained enterprise business status scoring model is used to predict and score the test set to obtain the business status score of the target enterprise to be verified.

[0124] Optionally, after determining the business status score output by the enterprise business status scoring model, the method may also include: the electronic device applies the business status score to government and enterprise marketing to assist offline contacts and online cross-marketing.

[0125] It should be noted that, at present, the expansion marketing of the government and enterprise market relies heavily on the manual street-sweeping investigation and inspection by front-line agents or account managers. Faced with a huge amount of corporate information, relying solely on the industrial and commercial enterprise status and manpower investment can no longer meet the market's demand for government and enterprise customer mining. After determining the business status score output by the enterprise business status scoring model, the electronic equipment can apply the business status score to government and enterprise marketing to assist offline contact and online cross-marketing. The whole process reduces the invalid enterprise contact rate while increasing the success rate of offline contact by account managers. Replacing manual labor with machines has formed a key breakthrough in enterprise precision marketing and has broad promotion prospects in the government and enterprise market.

[0126] In the embodiment of the present application, the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise are selected from the enterprise information integration library corresponding to the multi-source enterprise information; the feature engineering samples are constructed according to the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics; according to the feature engineering samples, the enterprise business status scoring model is constructed, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified. This method does not require human participation. Based on the enterprise information integration library corresponding to the multi-source enterprise information, a high-accuracy enterprise business status scoring model can be constructed, and then the business status score of the target enterprise to be verified can be automatically and accurately determined. The whole process improves the data processing efficiency while improving resource utilization.

[0127] The following describes a device for constructing a business operation status scoring model provided in an embodiment of the present application. The device for constructing a business operation status scoring model described below and the method for constructing a business operation status scoring model described above can refer to each other.

[0128] Figure 3 Schematic diagram of the structure of the device for constructing the enterprise business status scoring model provided in the embodiment of the present application. Figure 3 , an embodiment of the present application provides a device for constructing an enterprise business status scoring model, including the following feature determination module 301 and model construction module 302.

[0129] The feature determination module 301 is used to select the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information; and construct a feature engineering sample according to the business characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics; The model building module 302 is used to build an enterprise business status scoring model based on the feature engineering sample, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0130] Optionally, the model building module 302 is specifically used to obtain a preset enterprise, which contains labeled effective enterprise features; construct multiple base decision trees based on the feature engineering sample and the labeled effective enterprise features; and construct a business status scoring model for the enterprise based on the multiple base decision trees.

[0131] Optionally, the construction device also includes a data processing module, which is used to extract structural features of multi-source enterprise information to obtain multi-source enterprise structural information; determine enterprise features that match preset enterprises based on the multi-source enterprise structural information; and construct the enterprise information integration library based on the enterprise features.

[0132] Optionally, the preset enterprise includes at least one preset sub-enterprise, and the data processing module is specifically used to perform the following operations for any preset sub-enterprise of the at least one preset sub-enterprise: calculate the first Internet feature and the first hotline traffic feature of each enterprise structure information in the multi-source enterprise structure information; determine the second Internet feature with the highest similarity to the preset sub-enterprise from all the first Internet features, and determine the second hotline traffic feature with the highest similarity to the preset sub-enterprise from all the first hotline traffic features; determine the integrated quality feature of the preset sub-enterprise based on the second Internet feature and the second hotline traffic feature; and determine the second Internet feature, the second hotline traffic feature and the integrated quality feature as the enterprise characteristics of the preset sub-enterprise.

[0133] Optionally, the model building module 302 is specifically used to construct an original feature set based on the feature engineering sample and the labeled effective enterprise features; from the original feature set, a plurality of training subsets are obtained by random sampling with replacement, and the number of features in each training subset is the same as the number of features in the original feature set; and according to the multiple training subsets, corresponding base decision trees are constructed.

[0134] Optionally, the data processing module is specifically used to perform the following operations for each enterprise structure information: determine the first Internet feature based on the industrial and commercial enterprise name, enterprise identity, contact number, detailed address and geographic coordinate information in the enterprise structure information; determine the first hotline call feature based on the industrial and commercial enterprise name, the enterprise identity and the contact number.

[0135] The electronic device involved in the embodiments of the present application may be a device that provides voice and / or data connectivity to a user, a handheld device with a wireless connection function, or other processing devices connected to a wireless modem, etc. In different systems, the names of electronic devices may also be different. For example, in a 5G system, an electronic device may be called a user equipment (UE).

[0136] Figure 4 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, referring to Figure 4 An embodiment of the present application further provides an electronic device, which may include: a memory 410, a transceiver 420 and a processor 430.

[0137] The memory 410 is used to store computer programs; the transceiver 420 is used to send and receive data under the control of the processor 430; the processor 430 is used to read the computer program in the memory 410 and perform the following operations.

[0138] From the enterprise information integration library corresponding to the multi-source enterprise information, the industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise are selected; according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics, a feature engineering sample is constructed; according to the feature engineering sample, an enterprise business status scoring model is constructed, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0139] Among them, Figure 4 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 430 and various circuits of memory represented by memory 410 are linked together. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 420 may be a plurality of components, including a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. For different user devices, the user interface 440 may also be an interface capable of externally connecting to the required internal devices.

[0140] The processor 430 is responsible for managing the bus architecture and general processing, and the memory 410 can store data used by the processor 430 when performing operations.

[0141] The processor 430 is used to execute any of the methods provided in the embodiments of the present application according to the obtained executable instructions by calling the computer program stored in the memory 410. The processor and the memory may also be arranged physically separately.

[0142] It should be noted here that the electronic device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as those in the method embodiment will not be described in detail here.

[0143] Figure 5 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, such as Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the computer program in the memory 530 to execute the steps of the method for constructing the enterprise business status scoring model, including the following steps.

[0144] From the enterprise information integration library corresponding to the multi-source enterprise information, the industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise are selected; according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics, a feature engineering sample is constructed; according to the feature engineering sample, an enterprise business status scoring model is constructed, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0145] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0146] On the other hand, an embodiment of the present application also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the steps of the method for constructing an enterprise business status scoring model provided in the above embodiments, including the following steps.

[0147] From the enterprise information integration library corresponding to the multi-source enterprise information, the industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise are selected; according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics, a feature engineering sample is constructed; according to the feature engineering sample, an enterprise business status scoring model is constructed, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0148] On the other hand, an embodiment of the present application also provides a processor-readable storage medium, which stores a computer program. The computer program is used to enable the processor to execute the steps of the method for constructing an enterprise business status scoring model provided in the above embodiments, including the following steps.

[0149] From the enterprise information integration library corresponding to the multi-source enterprise information, the industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise are selected; according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics and the integrated quality characteristics, a feature engineering sample is constructed; according to the feature engineering sample, an enterprise business status scoring model is constructed, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

[0150] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.

[0151] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0152] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for constructing a business operation status scoring model, characterized in that: include: From the enterprise information integration database corresponding to multi-source enterprise information, select the enterprise's industrial and commercial characteristics, Internet characteristics, hotline traffic characteristics and integration quality characteristics; Constructing a feature engineering sample according to the industrial and commercial characteristics, the Internet characteristics, the hotline traffic characteristics, and the integrated quality characteristics; A business operation status scoring model is constructed based on the feature engineering samples, and the business operation status scoring model is used to score the business status of the target enterprise to be verified.

2. The method for constructing a business operation status scoring model according to claim 1, characterized in that: The step of constructing a business operation status scoring model based on the feature engineering sample includes: Acquire a preset enterprise, wherein the preset enterprise contains marked valid enterprise features; Constructing multiple base decision trees according to the feature engineering samples and the labeled effective enterprise features; The enterprise operation status scoring model is constructed based on the multiple base decision trees.

3. The method for constructing a business operation status scoring model according to claim 1, characterized in that: The method further comprises: Extracting structural features of multi-source enterprise information to obtain multi-source enterprise structural information; Determining enterprise characteristics matching a preset enterprise according to the multi-source enterprise structure information; According to the enterprise characteristics, the enterprise information integration library is constructed.

4. The method for constructing a business operation status scoring model according to claim 3, characterized in that: The preset enterprise includes at least one preset sub-enterprise, and determining enterprise characteristics matching the preset enterprise according to the multi-source enterprise structure information includes: The following operations are performed for any preset sub-enterprise of the at least one preset sub-enterprise: Calculating a first Internet feature and a first hotline traffic feature of each enterprise structure information in the multi-source enterprise structure information; Determine, from all first Internet features, a second Internet feature with the highest similarity to the preset sub-enterprise, and determine, from all first hotline traffic features, a second hotline traffic feature with the highest similarity to the preset sub-enterprise; Determining the integration quality characteristics of the preset subsidiary enterprise according to the second Internet characteristics and the second hotline traffic characteristics; The second Internet feature, the second hotline traffic feature and the integration quality feature are determined as enterprise features of the preset subsidiary enterprise.

5. The method for constructing a business operation status scoring model according to claim 2, characterized in that: The step of constructing a plurality of base decision trees according to the feature engineering samples and the annotated effective enterprise features includes: Constructing an original feature set according to the feature engineering sample and the annotated effective enterprise features; From the original feature set, a plurality of training subsets are obtained by random sampling with replacement, wherein the number of features in each training subset is the same as the number of features in the original feature set; According to the multiple training subsets, corresponding base decision trees are constructed.

6. The method for constructing a business operation status scoring model according to claim 4, characterized in that: The calculating of the first Internet feature and the first hotline traffic feature of each enterprise structure information in the multi-source enterprise structure information includes: The following operations are performed for each enterprise structure information: Determine the first Internet feature according to the business enterprise name, enterprise identity, contact number, detailed address and geographic coordinate information in the enterprise structure information; The first hotline traffic characteristics are determined according to the industrial and commercial enterprise name, the enterprise identity identifier and the contact number.

7. A device for constructing a business operation status scoring model, characterized in that: include: A feature determination module is used to select the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics of the enterprise from the enterprise information integration library corresponding to the multi-source enterprise information; and construct a feature engineering sample based on the business characteristics, Internet characteristics, hotline traffic characteristics and integrated quality characteristics; The model building module is used to build an enterprise business status scoring model based on the feature engineering sample, and the enterprise business status scoring model is used to score the business status of the target enterprise to be verified.

8. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the steps of the method for constructing the enterprise business status scoring model described in any one of claims 1 to 6 are implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for constructing an enterprise business status scoring model as described in any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for constructing an enterprise business status scoring model as described in any one of claims 1 to 6 are implemented.