WO2024066036A1 - 一种基于安全管理措施和保函的投保人推荐方法 - Google Patents

一种基于安全管理措施和保函的投保人推荐方法 Download PDF

Info

Publication number
WO2024066036A1
WO2024066036A1 PCT/CN2022/136946 CN2022136946W WO2024066036A1 WO 2024066036 A1 WO2024066036 A1 WO 2024066036A1 CN 2022136946 W CN2022136946 W CN 2022136946W WO 2024066036 A1 WO2024066036 A1 WO 2024066036A1
Authority
WO
WIPO (PCT)
Prior art keywords
guarantee
letter
insured
safety management
management measures
Prior art date
Application number
PCT/CN2022/136946
Other languages
English (en)
French (fr)
Inventor
吴承科
郭媛君
刘祥飞
杨之乐
冯伟
王尧
Original Assignee
深圳先进技术研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Publication of WO2024066036A1 publication Critical patent/WO2024066036A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Definitions

  • the present invention relates to the application field of letters of guarantee, and in particular to a method for recommending policyholders based on security management measures and letters of guarantee.
  • the electronic letter of guarantee for construction projects is an indispensable part of the construction engineering field. It refers to a written credit guarantee certificate issued by a bank, insurance company, or guarantee company to a third party at the request of an applicant. Letters of guarantee are mostly issued by insurance companies with strong financial strength. They can be based on credit and do not require traditional mortgage guarantees, thereby alleviating the financial pressure of enterprises in public resource transactions. Therefore, the letter of guarantee has a certain economic value.
  • the insured is recommended to the guarantor by evaluating the insured's letter of guarantee data. Since the letter of guarantee data cannot indicate the insured's risk, it is easy to recommend an insured with a high letter of guarantee value and high risk, resulting in the problem that the guarantor needs to bear too high a risk.
  • the technical problem to be solved by the present invention is that, in response to the above-mentioned defects of the prior art, a method for recommending policyholders based on security management measures and letters of guarantee is provided, aiming to solve the problem in the prior art of recommending policyholders to guarantors by evaluating the policyholders' letter of guarantee data. Since the letter of guarantee data cannot indicate the risks of the policyholders, it is easy to recommend policyholders with high letter of guarantee value and high risks, resulting in the guarantor having to bear excessive risks.
  • an embodiment of the present invention provides a method for recommending an insured based on security management measures and a letter of guarantee, wherein the method comprises:
  • a target policyholder is determined from among the policyholders according to the reliabilities respectively corresponding to the safety management measures.
  • obtaining the letter of guarantee characteristics corresponding to a plurality of policyholders respectively includes:
  • the letter of guarantee features corresponding to each of the policyholders are extracted from the letter of guarantee data.
  • the letter of guarantee feature is vector data
  • the insured persons are matched according to the letter of guarantee features to obtain a plurality of control groups, including:
  • the control groups are determined according to the similarities of the cosine vectors corresponding to the insured persons, wherein the similarity of the cosine vectors corresponding to the two insured persons in each control group is the highest.
  • the process of determining the insured value corresponding to each of the insured persons includes:
  • the guarantee value corresponding to the policyholder is obtained, and the insurance value corresponding to the policyholder is determined according to the difference between the guarantee value and the penalty value.
  • obtaining the value of the letter of guarantee corresponding to the policyholder includes:
  • the guarantee knowledge graph is input into a pre-trained target prediction model to obtain the guarantee value corresponding to the policyholder.
  • each of the control groups includes two policyholders
  • the intra-group comparison of the safety management measures and the insured values of the policyholders included in each of the control groups to obtain the reliability corresponding to each of the safety management measures includes:
  • a reliability reward is given to the safety management measure with the highest initial reliability in the control group to obtain the reliability corresponding to the safety management measure
  • the reliability penalty is performed on the safety management measure with the lowest initial reliability in the control group according to the reward and punishment value to obtain the reliability corresponding to the safety management measure.
  • determining a target policyholder from among the policyholders according to the reliability corresponding to each of the safety management measures includes:
  • the insured person corresponding to the safety management measure with the highest reliability is used as the target insured person.
  • an embodiment of the present invention further provides a policyholder recommendation system based on security management measures and a letter of guarantee, wherein the system comprises:
  • a matching module is used to obtain letter of guarantee characteristics corresponding to a number of policyholders, and match the policyholders according to the letter of guarantee characteristics to obtain a number of control groups;
  • An acquisition module used to acquire the safety management measures and insurance value corresponding to each of the insured persons, wherein the insurance value corresponding to each of the insured persons is determined based on the safety accident risk corresponding to the insured person and the value of the letter of guarantee;
  • a comparison module used for performing an intra-group comparison on the safety management measures and the insured values of the insured persons included in each of the control groups, and obtaining the reliability corresponding to each of the safety management measures;
  • the selection module is used to determine a target policyholder from among the policyholders according to the reliability corresponding to each of the safety management measures.
  • an embodiment of the present invention further provides a terminal, wherein the terminal includes a storage device and one or more processors; the storage device stores one or more programs; the program includes instructions for executing any of the above-mentioned methods for recommending policyholders based on security management measures and letters of guarantee; and the processor is used to execute the program.
  • an embodiment of the present invention also provides a computer-readable storage medium on which a plurality of instructions are stored, wherein the instructions are suitable for being loaded and executed by a processor to implement any of the steps of the above-mentioned method for recommending an insured based on security management measures and a letter of guarantee.
  • the embodiment of the present invention obtains the letter of guarantee characteristics corresponding to several policyholders, matches the policyholders according to the letter of guarantee characteristics, and obtains several control groups; obtains the safety management measures and insurance value corresponding to each policyholder, and the insurance value corresponding to each policyholder is determined based on the safety accident risk and the letter of guarantee value corresponding to the policyholder; performs intra-group comparison on the safety management measures and insurance values of the policyholders included in each control group to obtain the reliability corresponding to each safety management measure; and determines the target policyholder from the policyholders according to the reliability corresponding to each safety management measure.
  • the problem in the prior art of recommending policyholders to guarantors by evaluating the policyholder's letter of guarantee data is solved. Since the letter of guarantee data cannot indicate the risk of the policyholder, it is easy to recommend policyholders with high letter of guarantee value and high risk, resulting in the guarantor needing to bear excessive risk.
  • FIG1 is a flow chart of a method for recommending a policyholder based on security management measures and a letter of guarantee provided in an embodiment of the present invention.
  • FIG2 is a module diagram of a policyholder recommendation system based on security management measures and a letter of guarantee provided in an embodiment of the present invention.
  • FIG3 is a functional block diagram of a terminal provided by an embodiment of the present invention.
  • the present invention discloses a policyholder recommendation method based on security management measures and a letter of guarantee.
  • the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
  • the electronic letter of guarantee for construction projects is an indispensable part of the construction engineering field. It refers to a written credit guarantee certificate issued by a bank, insurance company, or guarantee company to a third party at the request of an applicant. Letters of guarantee are mostly issued by insurance companies with strong financial strength. They can be based on credit and do not require traditional mortgage guarantees, thereby alleviating the financial pressure of enterprises in public resource transactions. Therefore, the letter of guarantee has a certain economic value.
  • the insured is recommended to the guarantor by evaluating the insured's letter of guarantee data. Since the letter of guarantee data cannot indicate the insured's risk, it is easy to recommend an insured with a high letter of guarantee value and high risk, resulting in the problem that the guarantor needs to bear too high a risk.
  • the present invention provides a method for recommending an insured based on safety management measures and letters of guarantee, the method comprising: obtaining the letter of guarantee characteristics corresponding to a number of insureds, matching each of the insureds according to the letter of guarantee characteristics, and obtaining a number of control groups; obtaining the safety management measures and insurance values corresponding to each of the insureds, wherein the insurance value corresponding to each of the insureds is determined based on the safety accident risk and the letter of guarantee value corresponding to the insured; performing an intra-group comparison on the safety management measures and the insurance values of the insureds included in each of the control groups, and obtaining the reliability corresponding to each of the safety management measures; and determining the target insured from each of the insureds according to the reliability corresponding to each of the safety management measures.
  • the method solves the problem in the prior art of recommending insureds to guarantors by evaluating the insured's letter of guarantee data, and since the letter of guarantee data cannot indicate the insured's risk, it is easy to recommend insureds with high letter of guarantee value and high risk, resulting in the guarantor needing to bear excessive risk.
  • the method comprises:
  • Step S100 Obtain letter of guarantee characteristics corresponding to a number of policyholders, match the policyholders according to the letter of guarantee characteristics, and obtain a number of control groups.
  • this embodiment first determines multiple policyholders that the guarantor is interested in. For each policyholder, the corresponding guarantee data of the policyholder is obtained, and the guarantee characteristics of the policyholder are extracted based on the guarantee data. Then, based on the guarantee characteristics of each policyholder, policyholders with similar guarantee characteristics are divided into the same control group to obtain multiple control groups.
  • step S100 specifically includes:
  • Step S101 obtaining the guarantee letter data and the preset feature extraction template corresponding to each of the insured persons;
  • Step S102 extracting the letter of guarantee features corresponding to each of the insured persons from each of the letter of guarantee data according to the feature extraction template.
  • this embodiment pre-designs feature extraction templates for different types of policyholders, for example, policyholder types include contractors, suppliers, labor subcontractors, etc. Determine the feature extraction template corresponding to each current policyholder, and the feature extraction template contains the typical variable features of these policyholders, such as years of establishment, registered capital, revenue data, staff size, etc.
  • the feature extraction template can be used to extract the letter of guarantee features corresponding to each policyholder, and the letter of guarantee features of each policyholder can be uniformly represented in vector form.
  • the letter of guarantee feature is vector data
  • the insured persons are matched according to the letter of guarantee features to obtain a plurality of control groups, including:
  • Step S103 determining the cosine vector similarity between the insured persons according to the letter of guarantee characteristics of the insured persons;
  • Step S104 determining each control group according to the cosine vector similarities corresponding to each of the policyholders, wherein the cosine vector similarity corresponding to two policyholders in each control group is the highest.
  • this embodiment represents the guarantee characteristics of each policyholder in the form of a vector.
  • the cosine vector similarity is calculated for the guarantee characteristics of the policyholder and the guarantee characteristics of other policyholders in turn.
  • the higher the cosine vector similarity the closer the two vectors are, that is, the more similar the guarantee characteristics of the two policyholders are.
  • the other policyholder with the highest cosine vector similarity with the policyholder is assigned to the same control group. Since the guarantee characteristics of the two policyholders in the same control group are highly similar, the impact of other factors other than the guarantee characteristics on the value of the guarantee can be effectively observed through each control group.
  • the method further includes:
  • Step S200 Obtain the safety management measures and insurance value corresponding to each of the insured persons, wherein the insurance value corresponding to each of the insured persons is determined based on the safety accident risk corresponding to the insured person and the value of the letter of guarantee.
  • this embodiment not only needs to consider the insurance value of each policyholder when recommending policyholders, but also needs to comprehensively consider the safety management measures taken by each policyholder.
  • the insurance value of each policyholder is determined based on the corresponding safety accident risk and the value of the letter of guarantee of the policyholder, thereby excluding high-risk policyholders.
  • the process of determining the insurance value corresponding to each of the insured persons includes:
  • Step S201 Obtain the safety accident record corresponding to the policyholder, and determine the safety accident risk corresponding to the policyholder according to the safety accident record;
  • Step S202 determining a penalty value according to the safety accident risk corresponding to the policyholder
  • Step S203 Obtain the letter of guarantee value corresponding to the policyholder, and determine the insurance value corresponding to the policyholder according to the difference between the letter of guarantee value and the penalty value.
  • the safety accident record of the insured's company is obtained, and then the safety accident risk of the insured is predicted based on the safety accident record.
  • the insured's safety accident risk is high, it means that the insured is a high-risk insured, and a penalty value is assigned to the insured.
  • the penalty value is subtracted from the corresponding guarantee value to generate the final insurance value, so as to achieve the purpose of reducing the insurance value of high-risk insured and reduce the probability of recommending high-risk insured to the guarantor.
  • obtaining the value of the letter of guarantee corresponding to the policyholder includes:
  • Step S204 input the guarantee data corresponding to the policyholder into a pre-trained target extraction model to obtain a guarantee knowledge graph corresponding to the policyholder;
  • Step S205 input the letter of guarantee knowledge graph into a pre-trained target prediction model to obtain the letter of guarantee value corresponding to the policyholder.
  • this embodiment predicts the value of each policyholder's letter of guarantee by constructing a knowledge graph.
  • This embodiment pre-constructs a target extraction model, which has previously learned the complex mapping relationship between different letter of guarantee data and the letter of guarantee knowledge graph through massive data. After inputting the letter of guarantee data of each policyholder into the trained target extraction model, the target extraction model can automatically extract the information in the letter of guarantee data and generate the letter of guarantee knowledge graph corresponding to the policyholder.
  • This embodiment also pre-constructs a target prediction model, which has previously learned the complex mapping relationship between different letter of guarantee knowledge graphs and the letter of guarantee value through massive data. Therefore, the letter of guarantee knowledge graph of the policyholder is input into the trained target prediction model, and the target prediction model can output the letter of guarantee value corresponding to the policyholder.
  • the specific process of generating the guarantee value includes:
  • the letter of guarantee knowledge graph is input into the target prediction model to obtain the predicted letter of guarantee value.
  • the training process corresponding to the target extraction model includes:
  • Acquire historical letter of guarantee data and determine a number of first training data according to the historical letter of guarantee data, wherein each of the first training data includes a sentence in the historical letter of guarantee data and annotation information corresponding to the sentence, and the annotation information is used to reflect the entities included in the sentence and the relationship between the entities;
  • the input data of the target bidirectional language model is the guarantee data sentence masked by the mask
  • the output data of the target bidirectional language model is the predicted words masked by the mask
  • the target bidirectional language model is adjusted to obtain an extraction model, wherein the input data of the extraction model is the sentences in the historical letter of guarantee data, and the output data of the extraction model is the predicted entities contained in the sentences and the relationships between the entities;
  • the extraction model is iteratively trained according to each of the first training data to obtain the target extraction model.
  • the training process corresponding to the target extraction model includes:
  • Acquire historical letter of guarantee data and determine a number of first training data according to the historical letter of guarantee data, wherein each of the first training data includes a sentence in the historical letter of guarantee data and annotation information corresponding to the sentence, and the annotation information is used to reflect the entities included in the sentence and the relationship between the entities;
  • the input data of the target bidirectional language model is the guarantee data sentence masked by the mask
  • the output data of the target bidirectional language model is the predicted words masked by the mask
  • the target bidirectional language model is adjusted to obtain an extraction model, wherein the input data of the extraction model is the sentences in the historical letter of guarantee data, and the output data of the extraction model is the predicted entities contained in the sentences and the relationships between the entities;
  • the extraction model is iteratively trained according to each of the first training data to obtain the target extraction model.
  • the training process corresponding to the target bidirectional language model includes:
  • the updated bidirectional language model is used as the bidirectional language model, and the masking of words in the letter of guarantee data sentence is continued to obtain a masked sentence, until the first loss function value converges to the target value, thereby obtaining the target bidirectional language model.
  • masking the words in the letter of guarantee data sentence by masking to obtain the masked sentence includes:
  • the masking probability corresponding to each word in the letter of guarantee data sentence is proportional to the contribution of the word to the letter of guarantee value of the historical letter of guarantee data
  • the masked sentence is obtained by masking the words in the letter of guarantee data sentence based on the masking probabilities respectively corresponding to the words in the letter of guarantee data sentence through the mask.
  • generating a letter of guarantee knowledge graph corresponding to the letter of guarantee data according to each of the triples includes:
  • each node in the guarantee knowledge graph is connected to obtain the guarantee knowledge graph, wherein different types of relationships correspond to different types of connections.
  • the letter of guarantee knowledge graph further includes attention weights corresponding to each of the nodes, and the process of determining the attention weights corresponding to each of the nodes includes:
  • a neighborhood search is performed on the letter of guarantee knowledge graph to obtain all associated nodes corresponding to the node;
  • the attention weight corresponding to the node is determined according to the size of the relationship box.
  • the training process corresponding to the target prediction model includes:
  • Obtain a historical letter of guarantee knowledge graph and a letter of guarantee value corresponding to the historical letter of guarantee knowledge graph input the historical letter of guarantee knowledge graph into an untrained prediction model, and obtain a trained letter of guarantee value; wherein the letter of guarantee type corresponding to the historical letter of guarantee knowledge graph is the same as the letter of guarantee data, and the prediction model is a graph attention model;
  • the method further includes:
  • Step S300 perform an intra-group comparison on the safety management measures and the insured values of the policyholders included in each control group to obtain the reliability corresponding to each safety management measure.
  • this embodiment can accurately determine the reliability corresponding to each type of safety management measure by comparing the safety management measures and bid value of each insured person in each control group.
  • each of the control groups includes two policyholders
  • the intra-group comparison of the safety management measures and the insured values of the policyholders included in each of the control groups to obtain the reliability corresponding to each of the safety management measures includes:
  • Step S301 determining the initial reliability of the two safety management measures corresponding to each control group according to the two insured values corresponding to the control group;
  • Step S302 determining the reward or punishment value corresponding to each control group according to the difference between the two insured values corresponding to the control group;
  • Step S303 performing reliability reward on the safety management measure with the highest initial reliability in the control group according to the reward and punishment value, to obtain the reliability corresponding to the safety management measure;
  • Step S304 Perform reliability penalty on the safety management measure with the lowest initial reliability in the control group according to the reward and penalty value to obtain the reliability corresponding to the safety management measure.
  • the only variable that causes the different insurance values of the two insured persons is the different safety accident risks caused by the safety management measures adopted by the two insured persons.
  • this embodiment also sets a reward and punishment value for the deviation between the two insurance values, rewards the reliability of safety management measures with significantly good effects, and punishes the reliability of safety management measures with significantly poor effects, so as to better distinguish safety management measures with different effects.
  • the method further includes:
  • Step S400 determining a target policyholder from among the policyholders according to the reliabilities corresponding to the safety management measures.
  • the appropriate target insured can be recommended to the guarantor based on the reliability of each safety management measure.
  • step S400 specifically includes:
  • Step S401 The insured person corresponding to the safety management measure with the highest reliability is taken as the target insured person.
  • the policyholders whose safety management measures are most reliable are less likely to have safety accident risks, and therefore can be selected as target policyholders.
  • the guarantee value corresponding to each insured can also be obtained, and the insured with the highest guarantee value can be selected as candidate insureds, and then the insured with the highest reliability of safety management measures can be selected from the candidate insureds as the target insured. In this way, the target insured with high guarantee value and low safety accident risk can be accurately screened out.
  • the present invention further provides a policyholder recommendation system based on security management measures and a letter of guarantee, as shown in FIG2 , the system includes:
  • Matching module 01 is used to obtain letter of guarantee characteristics corresponding to a number of policyholders, and match the policyholders according to the letter of guarantee characteristics to obtain a number of control groups;
  • the acquisition module 02 is used to acquire the safety management measures and the insured value corresponding to each of the insured persons, wherein the insured value corresponding to each of the insured persons is determined based on the safety accident risk corresponding to the insured person and the value of the letter of guarantee;
  • Comparison module 03 used for performing intra-group comparison on the safety management measures and the insured values of the insured persons included in each control group, and obtaining the reliability corresponding to each safety management measure;
  • the selection module 04 is used to determine a target policyholder from among the policyholders according to the reliability corresponding to each of the safety management measures.
  • the present invention also provides a terminal, whose principle block diagram can be shown in Figure 3.
  • the terminal includes a processor, a memory, a network interface, and a display screen connected through a system bus.
  • the processor of the terminal is used to provide computing and control capabilities.
  • the memory of the terminal includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system and a computer program.
  • the internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium.
  • the network interface of the terminal is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a policyholder recommendation method based on security management measures and a letter of guarantee is implemented.
  • the display screen of the terminal can be a liquid crystal display or an electronic ink display.
  • FIG3 is only a block diagram of a partial structure related to the solution of the present invention, and does not constitute a limitation on the terminal to which the solution of the present invention is applied.
  • the specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
  • the terminal has one or more programs stored in its memory and is configured to be executed by one or more processors.
  • the one or more programs include instructions for performing a policyholder recommendation method based on security management measures and a letter of guarantee.
  • Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory.
  • Volatile memory can include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
  • the present invention discloses a method for recommending policyholders based on safety management measures and letters of guarantee, the method comprising: obtaining letter of guarantee characteristics corresponding to a number of policyholders, matching each policyholder according to the letter of guarantee characteristics, and obtaining a number of control groups; obtaining safety management measures and insurance values corresponding to each policyholder, wherein the insurance value corresponding to each policyholder is determined based on the safety accident risk and letter of guarantee value corresponding to the policyholder; performing intra-group comparison on the safety management measures and insurance values of the policyholders included in each control group, and obtaining the reliability corresponding to each safety management measure; and determining the target policyholder from each policyholder according to the reliability corresponding to each safety management measure.
  • the method solves the problem in the prior art of recommending policyholders to guarantors by evaluating the policyholder's letter of guarantee data, and since the letter of guarantee data cannot indicate the risk of the policyholder, it is easy to recommend policyholders with high letter of guarantee value and high risk, resulting in the guarantor needing to bear excessive risk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Technology Law (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

本发明公开了一种基于安全管理措施和保函的投保人推荐方法，通过获取若干投保人分别对应的保函特征，根据各保函特征对各投保人进行匹配，得到若干对照组；获取各投保人分别对应的安全管理措施和投保价值，每一投保人对应的投保价值基于该投保人对应的安全事故风险和保函价值确定；对各对照组中包含的投保人的安全管理措施和投保价值进行组内比对，得到各安全管理措施分别对应的可靠度；根据各安全管理措施分别对应的可靠度，从各投保人中确定目标投保人。解决了现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。

Description

一种基于安全管理措施和保函的投保人推荐方法 技术领域
本发明涉及保函应用领域，尤其涉及的是一种基于安全管理措施和保函的投保人推荐方法。
 背景技术
建设工程电子保函为建设工程领域不可或缺的一环，其指的是银行、保险公司、担保公司应申请人的请求，向第三方开具的一种书面信用担保凭证。保函多由具有强大资金实力的保险公司出具，可以以信用为基础，无需传统抵质押担保，从而缓解企业在公共资源交易中的资金压力。因此保函具有一定的经济价值。现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
因此，现有技术还有待改进和发展。
 技术问题
本发明要解决的技术问题在于，针对现有技术的上述缺陷，提供一种基于安全管理措施和保函的投保人推荐方法，旨在解决现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
 技术解决方案
本发明解决问题所采用的技术方案如下：
第一方面，本发明实施例提供一种基于安全管理措施和保函的投保人推荐方法，其中，所述方法包括：
获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；
获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；
对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；
根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
在一种实施方式中，所述获取若干投保人分别对应的保函特征，包括：
获取各所述投保人分别对应的保函数据和预设的特征提取模板；
根据所述特征提取模板，从各所述保函数据中提取各所述投保人分别对应的所述保函特征。
在一种实施方式中，所述保函特征为向量数据，所述根据各所述保函特征对各所述投保人进行匹配，得到若干对照组，包括：
根据各所述投保人分别的所述保函特征，确定各所述投保人相互对应的余弦向量相似度；
根据各所述投保人相互对应的所述余弦向量相似度，确定各所述对照组，其中，每一所述对照组中的两个所述投保人相互对应的所述余弦向量相似度最高。
在一种实施方式中，每一所述投保人对应的投保价值的确定过程包括：
获取该投保人对应的安全事故记录，根据所述安全事故记录确定该投保人对应的所述安全事故风险；
根据该投保人对应的所述安全事故风险，确定惩罚值；
获取该投保人对应的所述保函价值，根据所述保函价值和所述惩罚值的差值确定该投保人对应的所述投保价值。
在一种实施方式中，所述获取该投保人对应的所述保函价值，包括：
将该投保人对应的所述保函数据输入预先经过训练的目标提取模型，得到该投保人对应的保函知识图谱；
将所述保函知识图谱输入预先经过训练的目标预测模型，得到该投保人对应的所述保函价值。
在一种实施方式中，每一所述对照组包括两个所述投保人，所述对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度，包括：
根据每一所述对照组对应的两个所述投保价值，确定该对照组对应的两个所述安全管理措施的初始可靠度；
根据每一所述对照组对应的两个所述投保价值的差值，确定该对照组对应的奖惩值；
根据所述奖惩值对该对照组中所述初始可靠度最高的所述安全管理措施进行可靠度奖励，得到该安全管理措施对应的所述可靠度；
根据所述奖惩值对该对照组中所述初始可靠度最低的所述安全管理措施进行可靠度惩罚，得到该安全管理措施对应的所述可靠度。
在一种实施方式中，所述根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人，包括：
将所述可靠度最高的所述安全管理措施对应的所述投保人作为所述目标投保人。
第二方面，本发明实施例还提供一种基于安全管理措施和保函的投保人推荐系统，其中，所述系统包括：
匹配模块，用于获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；
获取模块，用于获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；
比对模块，用于对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；
选择模块，用于根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
第三方面，本发明实施例还提供一种终端，其中，所述终端包括有存储所器和一个以上处理器；所述存储器存储有一个以上的程序；所述程序包含用于执行如上述任一所述的基于安全管理措施和保函的投保人推荐方法的指令；所述处理器用于执行所述程序。
第四方面，本发明实施例还提供计算机可读存储介质，其上存储有多条指令，其中，所述指令适用于由处理器加载并执行，以实现上述任一述的基于安全管理措施和保函的投保人推荐方法的步骤。
 有益效果
本发明的有益效果：本发明实施例通过获取若干投保人分别对应的保函特征，根据各保函特征对各投保人进行匹配，得到若干对照组；获取各投保人分别对应的安全管理措施和投保价值，每一投保人对应的投保价值基于该投保人对应的安全事故风险和保函价值确定；对各对照组中包含的投保人的安全管理措施和投保价值进行组内比对，得到各安全管理措施分别对应的可靠度；根据各安全管理措施分别对应的可靠度，从各投保人中确定目标投保人。解决了现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
 附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明中记载的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的基于安全管理措施和保函的投保人推荐方法的流程示意图。
图2是本发明实施例提供的基于安全管理措施和保函的投保人推荐系统的模块示意图。
图3是本发明实施例提供的终端的原理框图。
 本发明的实施方式
本发明公开了一种基于安全管理措施和保函的投保人推荐方法，为使本发明的目的、技术方案及效果更加清楚、明确，以下参照附图并举实施例对本发明进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本发明，并不用于限定本发明。
本技术领域技术人员可以理解，除非特意声明，这里使用的单数形式“一”、“一个”、“所述”和“该”也可包括复数形式。应该进一步理解的是，本发明的说明书中使用的措辞“包括”是指存在所述特征、整数、步骤、操作、元件和/或组件，但是并不排除存在或添加一个或多个其他特征、整数、步骤、操作、元件、组件和/或它们的组。 应该理解，当我们称元件被“连接”或“耦接”到另一元件时，它可以直接连接或耦接到其他元件，或者也可以存在中间元件。此外，这里使用的“连接”或“耦接”可以包括无线连接或无线耦接。这里使用的措辞“和/或”包括一个或更多个相关联的列出项的全部或任一单元和全部组合。
本技术领域技术人员可以理解，除非另外定义，这里使用的所有术语(包括技术术语和科学术语)，具有与本发明所属领域中的普通技术人员的一般理解相同的意义。还应该理解的是，诸如通用字典中定义的那些术语，应该被理解为具有与现有技术的上下文中的意义一致的意义，并且除非像这里一样被特定定义，否则不会用理想化或过于正式的含义来解释。
建设工程电子保函为建设工程领域不可或缺的一环，其指的是银行、保险公司、担保公司应申请人的请求，向第三方开具的一种书面信用担保凭证。保函多由具有强大资金实力的保险公司出具，可以以信用为基础，无需传统抵质押担保，从而缓解企业在公共资源交易中的资金压力。因此保函具有一定的经济价值。现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
针对现有技术的上述缺陷，本发明提供一种基于安全管理措施和保函的投保人推荐方法，所述方法包括：获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。解决了现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
示例性方法
如图1所示，所述方法包括：
步骤S100、获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组。
具体地，本实施例首先确定担保人感兴趣的多个投保人。针对每一投保人，获取该投保人对应的保函数据，根据保函数据提取该投保人的保函特征。然后根据各投保人的保函特征，将保函特征相似的各投保人分至同一对照组，得到多个对照组。
在一种实现方式中，所述步骤S100具体包括：
步骤S101、获取各所述投保人分别对应的保函数据和预设的特征提取模板；
步骤S102、根据所述特征提取模板，从各所述保函数据中提取各所述投保人分别对应的所述保函特征。
具体地，为了提取保函特征，本实施例预先设计了面向不同类型的投保人的特征提取模板，例如投保人类型包括承包商、供应商、劳务分包等。确定当前的各投保人对应的特征提取模板，该特征提取模板包含这些投保人的典型变量特征，例如成立年限、注册资本、营收数据、人员规模等。通过获取各投保人的保函数据，再使用该特征提取模板即可提取出各投保人分别对应的保函特征，并统一以向量形式表示各投保人的保函特征。
在一种实现方式中，所述保函特征为向量数据，所述根据各所述保函特征对各所述投保人进行匹配，得到若干对照组，包括：
步骤S103、根据各所述投保人分别的所述保函特征，确定各所述投保人相互对应的余弦向量相似度；
步骤S104、根据各所述投保人相互对应的所述余弦向量相似度，确定各所述对照组，其中，每一所述对照组中的两个所述投保人相互对应的所述余弦向量相似度最高。
具体地，本实施例以向量形式表示各投保人的保函特征。针对每一投保人，将该投保人的保函特征与其他人的保函特征依次计算余弦向量相似度，余弦向量相似度越高，表示两个向量越接近，即两个投保人的保函特征越相似。将与该投保人与余弦向量相似度最高的另一投保人分至同一对照组。由于位于同一对照组的两个投保人之间的保函特征高度相似，因此可以通过各对照组有效观测到除保函特征之外的其他因素对保函价值的影响。
如图1所示，所述方法还包括：
步骤S200、获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定。
具体地，为了向担保人推荐更优质的投保人，本实施例在进行投保人推荐的时候不仅需要考虑各投保人的投保价值，还需要综合考虑到各投保人分别采取的安全管理措施。在一种实现方式中，每一投保人的投保价值均基于该投保人对应的安全事故风险和保函价值确定，从而排除高危投保人。
在一种实现方式中，每一所述投保人对应的投保价值的确定过程包括：
步骤S201、获取该投保人对应的安全事故记录，根据所述安全事故记录确定该投保人对应的所述安全事故风险；
步骤S202、根据该投保人对应的所述安全事故风险，确定惩罚值；
步骤S203、获取该投保人对应的所述保函价值，根据所述保函价值和所述惩罚值的差值确定该投保人对应的所述投保价值。
具体地，针对每一投保人，获取该投保人的公司的安全事故记录，然后根据安全事故记录预测出该投保人的安全事故风险。当该投保人的安全事故风险较高时，表示该投保人为高危投保人，则对其赋予惩罚值，在其对应的保函价值的基础上减去惩罚值生成最终的投保价值，以达到降低高危投保人的投保价值的目的，减少为担保人推荐高危投保人的概率。
在一种实现方式中，所述获取该投保人对应的所述保函价值，包括：
步骤S204、将该投保人对应的所述保函数据输入预先经过训练的目标提取模型，得到该投保人对应的保函知识图谱；
步骤S205、将所述保函知识图谱输入预先经过训练的目标预测模型，得到该投保人对应的所述保函价值。
具体地，本实施例是通过构建知识图谱的方式来预测每一投保人的保函价值。本实施例预先构建了一个目标提取模型，该目标提取模型预先通过海量数据学习了不同保函数据与保函知识图谱之间的复杂映射关系。将每一投保人的保函数据输入已训练的目标提取模型后，该目标提取模型即可自动提取保函数据中的信息并生成该投保人对应的保函知识图谱。本实施例还预先构建了一个目标预测模型，该目标预测模型预先通过海量数据学习了不同保函知识图谱与保函价值之间的复杂映射关系，因此将该投保人的保函知识图谱输入已训练的目标预测模型，该目标预测模型即可输出该投保人对应的保函价值。
在一种实现方式中，所述保函价值的具体生成过程包括：
将保函数据输入已训练的目标提取模型，得到所述保函数据对应的若干三元组，每一所述三元组用于反映所述保函数据中两个实体之间的关系；
根据各所述三元组，生成所述保函数据对应的保函知识图谱；
确定所述保函数据对应的保函类型，根据所述保函类型确定所述保函数据对应的已训练的目标预测模型；
将所述保函知识图谱输入所述目标预测模型，得到预测保函价值。
在一种实现方式中，所述目标提取模型对应的训练过程包括：
获取历史保函数据，根据所述历史保函数据确定若干第一训练数据，其中，每一所述第一训练数据包含所述历史保函数据中的语句和该语句对应的标注信息，所述标注信息用于反映该语句中包含的实体和各实体之间的关系；
获取预先经过训练的目标双向语言模型，其中，所述目标双向语言模型的输入数据为通过掩码掩盖后的保函数据语句，所述目标双向语言模型的输出数据为预测的被所述掩码掩盖的词语；
对所述目标双向语言模型进行调整，得到提取模型，其中，所述提取模型的输入数据为所述历史保函数据中的语句，所述提取模型的输出数据为预测的该语句中包含的实体和各实体之间的关系；
根据各所述第一训练数据对所述提取模型进行迭代训练，得到所述目标提取模型。
在一种实现方式中，所述目标提取模型对应的训练过程包括：
获取历史保函数据，根据所述历史保函数据确定若干第一训练数据，其中，每一所述第一训练数据包含所述历史保函数据中的语句和该语句对应的标注信息，所述标注信息用于反映该语句中包含的实体和各实体之间的关系；
获取预先经过训练的目标双向语言模型，其中，所述目标双向语言模型的输入数据为通过掩码掩盖后的保函数据语句，所述目标双向语言模型的输出数据为预测的被所述掩码掩盖的词语；
对所述目标双向语言模型进行调整，得到提取模型，其中，所述提取模型的输入数据为所述历史保函数据中的语句，所述提取模型的输出数据为预测的该语句中包含的实体和各实体之间的关系；
根据各所述第一训练数据对所述提取模型进行迭代训练，得到所述目标提取模型。
在一种实现方式中，所述目标双向语言模型对应的训练过程包括：
根据所述历史保函数据，确定若干所述保函数据语句；
通过掩码对所述保函数据语句中的词语进行掩盖得到掩盖语句，根据被掩盖的词语生成所述掩盖语句对应的标签信息；
将所述掩盖语句输入未完成训练的双向语言模型，得到所述双向语言模型基于所述掩盖语句生成的预测词语；
根据所述预测词语和所述标签信息，生成所述双向语言模型对应的第一损失函数值；
判断所述第一损失函数值是否收敛至目标值，若否，根据所述第一损失函数值对所述双向语言模型进行参数更新，得到更新双向语言模型；
将所述更新双向语言模型作为所述双向语言模型，继续执行通过掩码对所述保函数据语句中的词语进行掩盖得到掩盖语句，直至所述第一损失函数值收敛至所述目标值，得到所述目标双向语言模型。
在一种实现方式中，所述通过掩码对所述保函数据语句中的词语进行掩盖得到掩盖语句，包括：
判断前一轮训练对应的所述第一损失函数值是否收敛至中间值，其中，所述中间值为首轮训练对应的所述第一损失函数值与所述目标值的中间数值；
若前一轮训练对应的所述第一损失函数值未收敛至所述中间值，通过所述掩码对所述保函数据语句中的词语进行随机掩盖得到所述掩盖语句；
若前一轮训练对应的所述第一损失函数值收敛至所述中间值时，获取所述保函数据语句中各词语分别对应的掩盖概率，其中，每一词语对应的所述掩盖概率与该词语对所述历史保函数据的保函价值的贡献度成正比关系；
通过所述掩码基于所述保函数据语句中各词语分别对应的所述掩盖概率，对所述保函数据语句中的词语进行掩盖得到所述掩盖语句。
在一种实现方式中，所述根据各所述三元组，生成所述保函数据对应的保函知识图谱，包括：
根据各所述三元组中包含的实体一一对应地生成所述保函知识图谱中的节点；
根据各所述三元组中包含的实体之间的关系，对所述保函知识图谱中的各节点进行连线，得到所述保函知识图谱，其中，不同类型的关系分别对应不同类型的连线。
在一种实现方式中，所述保函知识图谱还包括各所述节点分别对应的注意力权重，每一所述节点对应的所述注意力权重的确定过程包括：
以该节点为起始点，对所述保函知识图谱进行邻域搜索得到该节点对应的所有关联节点；
根据各所述关联节点，确定该节点对应的关系框，其中，所述关系框为包含有各所述关联节点的最小包围框；
根据所述关系框的大小，确定该节点对应的所述注意力权重。
在一种实现方式中，所述目标预测模型对应的训练过程包括：
获取历史保函知识图谱和所述历史保函知识图谱对应的保函价值，将所述历史保函知识图谱输入未完成训练的预测模型，得到训练保函价值；其中，所述历史保函知识图谱对应的保函类型与所述保函数据相同，所述预测模型为图注意力模型；
根据所述保函价值和所述训练保函价值的最小均方误差，确定所述预测模型对应的第二损失函数值；
判断所述第二损失函数值是否收敛至预设值，若否，根据所述第二损失函数值对所述预测模型进行参数更新，得到更新预测模型；
将所述更新预测模型作为所述预测模型，继续执行获取历史保函知识图谱和所述历史保函知识图谱对应的保函价值，将所述历史保函知识图谱输入未完成训练的预测模型的步骤，直至得到的所述第二损失函数值收敛至所述预设值，得到所述目标预测模型。
如图1所示，所述方法还包括：
步骤S300、对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度。
具体地，由于位于同一对照组中的各投保人的保函特征相似，因此导致同一对照组中的各投保人的投保价值出现差异的原因只能是安全管理措施的差别。所以本实施例通过对每一对照组中的各投保人的安全管理措施和投标价值进行比对，可以准确地确定各类安全管理措施分别对应的可靠度。
在一种实现方式中，每一所述对照组包括两个所述投保人，所述对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度，包括：
步骤S301、根据每一所述对照组对应的两个所述投保价值，确定该对照组对应的两个所述安全管理措施的初始可靠度；
步骤S302、根据每一所述对照组对应的两个所述投保价值的差值，确定该对照组对应的奖惩值；
步骤S303、根据所述奖惩值对该对照组中所述初始可靠度最高的所述安全管理措施进行可靠度奖励，得到该安全管理措施对应的所述可靠度；
步骤S304、根据所述奖惩值对该对照组中所述初始可靠度最低的所述安全管理措施进行可靠度惩罚，得到该安全管理措施对应的所述可靠度。
具体地，由于位于同一对照组的两个投保人的保函特征相似，因此导致这两个投保人的投保价值不同的唯一变量就是两个投保人分别采用的安全管理措施所导致的不同的安全事故风险。对于同一对照组中而言，投保价值越高，表示对应的投保人的安全事故风险可能越低，保函价值可能越高，则其采用的安全管理措施越可靠，则给予其较高的初始可靠度；投保价值越低，表示对应的投保人的安全事故风险可能越高，保函价值可能越低，则其采用的安全管理措施越不可靠，则给予其较低的初始可靠度。此外，同一对照组的两个投保价值之间的偏差越大，表示对应的两个投保人分别采用的安全管理措施的效果有显著差别，因此本实施例还会针对两个投保价值之间的偏差设定奖惩值，对效果显著好的安全管理措施的可靠度进行奖励，并对效果显著差的安全管理措施的可靠度进行惩罚，从而更好地区分出不同效果的安全管理措施。
如图1所示，所述方法还包括：
步骤S400、根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
具体地，由于安全管理措施的可靠度越高，表示该投保人的未来可能出现安全事故的风险越低，因此可以根据各安全管理措施的可靠度为担保人推荐合适的目标投保人。
在一种实现方式中，所述步骤S400具体包括：
步骤S401、将所述可靠度最高的所述安全管理措施对应的所述投保人作为所述目标投保人。
具体地，安全管理措施的可靠度最高的投保人，其出现安全事故风险的可能性较低，因此可以将其作为目标投保人。
在一种实现方式中，还可以获取各投保人分别对应的保函价值，将保函价值位于前若干位的投保人作为候选投保人，再从候选投保人中选择安全管理措施的可靠度最高的作为目标投保人。这样一来，就可以准确筛选出保函价值高，且安全事故风险低的目标投保人。
基于上述实施例，本发明还提供了一种基于安全管理措施和保函的投保人推荐系统，如图2所示，所述系统包括：
匹配模块01，用于获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；
获取模块02，用于获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；
比对模块03，用于对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；
选择模块04，用于根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
基于上述实施例，本发明还提供了一种终端，其原理框图可以如图3所示。该终端包括通过系统总线连接的处理器、存储器、网络接口、显示屏。其中，该终端的处理器用于提供计算和控制能力。该终端的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统和计算机程序。该内存储器为非易失性存储介质中的操作系统和计算机程序的运行提供环境。该终端的网络接口用于与外部的终端通过网络连接通信。该计算机程序被处理器执行时以实现基于安全管理措施和保函的投保人推荐方法。该终端的显示屏可以是液晶显示屏或者电子墨水显示屏。
本领域技术人员可以理解，图3中示出的原理框图，仅仅是与本发明方案相关的部分结构的框图，并不构成对本发明方案所应用于其上的终端的限定，具体的终端可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
在一种实现方式中，所述终端的存储器中存储有一个以上的程序，且经配置以由一个以上处理器执行所述一个以上程序包含用于进行基于安全管理措施和保函的投保人推荐方法的指令。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储于一非易失性计算机可读取存储介质中，该计算机程序在执行时，可包括如上述各方法的实施例的流程。其中，本发明所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器（ROM）、可编程ROM（PROM）、电可编程ROM（EPROM）、电可擦除可编程ROM（EEPROM）或闪存。易失性存储器可包括随机存取存储器（RAM）或者外部高速缓冲存储器。作为说明而非局限，RAM以多种形式可得，诸如静态RAM（SRAM）、动态RAM（DRAM）、同步DRAM（SDRAM）、双数据率SDRAM（DDRSDRAM）、增强型SDRAM（ESDRAM）、同步链路（Synchlink） DRAM（SLDRAM）、存储器总线（Rambus）直接RAM（RDRAM）、直接存储器总线动态RAM（DRDRAM）、以及存储器总线动态RAM（RDRAM）等。
综上所述，本发明公开了一种基于安全管理措施和保函的投保人推荐方法，所述方法包括：获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。解决了现有技术中通过评估投保人的保函数据向担保人推荐投保人，由于保函数据无法提示投保人的风险，因此容易推荐保函价值高，且风险也高的投保人，导致担保人需要承担过高的风险的问题。
应当理解的是，本发明的应用不限于上述的举例，对本领域普通技术人员来说，可以根据上述说明加以改进或变换，所有这些改进和变换都应属于本发明所附权利要求的保护范围。

Claims (10)

  1. 一种基于安全管理措施和保函的投保人推荐方法，其特征在于，所述方法包括：
    获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；
    获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；
    对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；
    根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
  2. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，所述获取若干投保人分别对应的保函特征，包括：
    获取各所述投保人分别对应的保函数据和预设的特征提取模板；
    根据所述特征提取模板，从各所述保函数据中提取各所述投保人分别对应的所述保函特征。
  3. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，所述保函特征为向量数据，所述根据各所述保函特征对各所述投保人进行匹配，得到若干对照组，包括：
    根据各所述投保人分别的所述保函特征，确定各所述投保人相互对应的余弦向量相似度；
    根据各所述投保人相互对应的所述余弦向量相似度，确定各所述对照组，其中，每一所述对照组中的两个所述投保人相互对应的所述余弦向量相似度最高。
  4. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，每一所述投保人对应的投保价值的确定过程包括：
    获取该投保人对应的安全事故记录，根据所述安全事故记录确定该投保人对应的所述安全事故风险；
    根据该投保人对应的所述安全事故风险，确定惩罚值；
    获取该投保人对应的所述保函价值，根据所述保函价值和所述惩罚值的差值确定该投保人对应的所述投保价值。
  5. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，所述获取该投保人对应的所述保函价值，包括：
    将该投保人对应的所述保函数据输入预先经过训练的目标提取模型，得到该投保人对应的保函知识图谱；
    将所述保函知识图谱输入预先经过训练的目标预测模型，得到该投保人对应的所述保函价值。
  6. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，每一所述对照组包括两个所述投保人，所述对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度，包括：
    根据每一所述对照组对应的两个所述投保价值，确定该对照组对应的两个所述安全管理措施的初始可靠度；
    根据每一所述对照组对应的两个所述投保价值的差值，确定该对照组对应的奖惩值；
    根据所述奖惩值对该对照组中所述初始可靠度最高的所述安全管理措施进行可靠度奖励，得到该安全管理措施对应的所述可靠度；
    根据所述奖惩值对该对照组中所述初始可靠度最低的所述安全管理措施进行可靠度惩罚，得到该安全管理措施对应的所述可靠度。
  7. 根据权利要求1所述的基于安全管理措施和保函的投保人推荐方法，其特征在于，所述根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人，包括：
    将所述可靠度最高的所述安全管理措施对应的所述投保人作为所述目标投保人。
  8. 一种基于安全管理措施和保函的投保人推荐系统，其特征在于，所述系统包括：
    匹配模块，用于获取若干投保人分别对应的保函特征，根据各所述保函特征对各所述投保人进行匹配，得到若干对照组；
    获取模块，用于获取各所述投保人分别对应的安全管理措施和投保价值，其中，每一所述投保人对应的所述投保价值基于该投保人对应的安全事故风险和保函价值确定；
    比对模块，用于对各所述对照组中包含的所述投保人的所述安全管理措施和所述投保价值进行组内比对，得到各所述安全管理措施分别对应的可靠度；
    选择模块，用于根据各所述安全管理措施分别对应的所述可靠度，从各所述投保人中确定目标投保人。
  9. 一种终端，其特征在于，所述终端包括有存储所器和一个以上处理器；所述存储器存储有一个以上的程序；所述程序包含用于执行如权利要求1-7中任一所述的基于安全管理措施和保函的投保人推荐方法的指令；所述处理器用于执行所述程序。
  10. 一种计算机可读存储介质，其上存储有多条指令，其特征在于，所述指令适用于由处理器加载并执行，以实现上述权利要求1-7任一述的基于安全管理措施和保函的投保人推荐方法的步骤。
PCT/CN2022/136946 2022-09-27 2022-12-06 一种基于安全管理措施和保函的投保人推荐方法 WO2024066036A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211181921.3 2022-09-27
CN202211181921.3A CN115760437A (zh) 2022-09-27 2022-09-27 一种基于安全管理措施和保函的投保人推荐方法

Publications (1)

Publication Number Publication Date
WO2024066036A1 true WO2024066036A1 (zh) 2024-04-04

Family

ID=85352005

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/136946 WO2024066036A1 (zh) 2022-09-27 2022-12-06 一种基于安全管理措施和保函的投保人推荐方法

Country Status (2)

Country Link
CN (1) CN115760437A (zh)
WO (1) WO2024066036A1 (zh)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8095396B1 (en) * 2008-03-27 2012-01-10 Asterisk Financial Group, Inc. Computer system for underwriting a personal guaranty liability by utilizing a risk apportionment system
CN109377388A (zh) * 2018-09-13 2019-02-22 平安医疗健康管理股份有限公司 医保投保方法、装置、计算机设备和存储介质
CN109670974A (zh) * 2018-12-14 2019-04-23 中国平安人寿保险股份有限公司 一种风险监控方法及装置、电子终端
CN110647809A (zh) * 2019-08-15 2020-01-03 中国平安人寿保险股份有限公司 基于图像分析的ai核保系统、方法及计算机可读存储介质
CN112182596A (zh) * 2020-03-24 2021-01-05 工保科技(浙江)有限公司 通过电子保函系统投标和开标的方法及数据保密方法
CN113034267A (zh) * 2021-03-26 2021-06-25 广东中盈盛达数字科技有限公司 一种基于供应链的电子保函系统
CN114663064A (zh) * 2022-03-30 2022-06-24 深圳市融锋科技有限公司 一种投标电子保函业务综合化智能管理系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8095396B1 (en) * 2008-03-27 2012-01-10 Asterisk Financial Group, Inc. Computer system for underwriting a personal guaranty liability by utilizing a risk apportionment system
CN109377388A (zh) * 2018-09-13 2019-02-22 平安医疗健康管理股份有限公司 医保投保方法、装置、计算机设备和存储介质
CN109670974A (zh) * 2018-12-14 2019-04-23 中国平安人寿保险股份有限公司 一种风险监控方法及装置、电子终端
CN110647809A (zh) * 2019-08-15 2020-01-03 中国平安人寿保险股份有限公司 基于图像分析的ai核保系统、方法及计算机可读存储介质
CN112182596A (zh) * 2020-03-24 2021-01-05 工保科技(浙江)有限公司 通过电子保函系统投标和开标的方法及数据保密方法
CN113034267A (zh) * 2021-03-26 2021-06-25 广东中盈盛达数字科技有限公司 一种基于供应链的电子保函系统
CN114663064A (zh) * 2022-03-30 2022-06-24 深圳市融锋科技有限公司 一种投标电子保函业务综合化智能管理系统

Also Published As

Publication number Publication date
CN115760437A (zh) 2023-03-07

Similar Documents

Publication Publication Date Title
US20210257066A1 (en) Machine learning based medical data classification method, computer device, and non-transitory computer-readable storage medium
CN111369342B (zh) 基于机器学习的贷款审批方法、装置、设备和存储介质
US10992763B2 (en) Dynamic interaction optimization and cross channel profile determination through online machine learning
US11449947B2 (en) Subrogation case management
US20170323272A1 (en) System environment for user-specific program aggregation and non-collocated third party system extraction and deployment
WO2021114620A1 (zh) 病历质控方法、装置、计算机设备和存储介质
US11164564B2 (en) Augmented intent and entity extraction using pattern recognition interstitial regular expressions
US11410448B2 (en) Predictive analysis systems and methods using machine learning
CN112016318B (zh) 基于解释模型的分诊信息推荐方法、装置、设备及介质
CN110377631B (zh) 案件信息处理方法、装置、计算机设备和存储介质
WO2019227576A1 (zh) 发票校验方法、装置、计算机设备及存储介质
CN112231224A (zh) 基于人工智能的业务系统测试方法、装置、设备和介质
CN112712429A (zh) 汇款业务审核方法、装置、计算机设备和存储介质
CN112036172B (zh) 基于模型的缩写数据的实体识别方法、装置和计算机设备
CN112908473A (zh) 基于模型的数据处理方法、装置、计算机设备和存储介质
CA3169288A1 (en) Knowledge graph based reasoning recommendation system and method
US11055200B2 (en) Systems and methods for validating domain specific models
CN112231556A (zh) 基于对话场景的用户画像方法、装置、设备及介质
WO2024066036A1 (zh) 一种基于安全管理措施和保函的投保人推荐方法
US20230112990A1 (en) Classification system
CN111723870A (zh) 基于人工智能的数据集获取方法、装置、设备和介质
CN112364136B (zh) 关键词生成方法、装置、设备及存储介质
CN113268563B (zh) 基于图神经网络的语义召回方法、装置、设备及介质
CN115130545A (zh) 数据处理方法、电子设备、程序产品及介质
CN110780850B (zh) 需求用例辅助生成方法、装置、计算机设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22960631

Country of ref document: EP

Kind code of ref document: A1