Integrated electronic quality inspection management platform for insurance policy

Abstract

The invention discloses an insurance policy comprehensive electronic quality inspection management platform, and relates to the technical field of data processing. The platform comprises a database construction module which obtains insurance policy data, preprocesses the insurance policy data and constructs an insurance policy database; the data quality inspection module is used for creating a quality inspection rule base and a verification rule base, acquiring a data analysis instruction and calling insurance policy data for verification and quality inspection; the risk identification module is used for establishing a risk identification model and carrying out risk assessment; the quality inspection evaluation module is used for recording and analyzing quality indexes during data quality inspection and evaluating the quality inspection quality; the process monitoring module is used for carrying out abnormity monitoring and early warning prompting on the quality inspection process; and the user management module is used for setting an operation access authority according to the identity of the user and carrying out data classification and encryption and user access control. According to the method, the insurance policy processing efficiency and accuracy and the security of the insurance policy data are improved by using automation and intelligence on the whole, the potential risk is reduced, and powerful support is provided for the safe and robust development of the insurance industry.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a comprehensive electronic quality inspection management platform for insurance policies. Background Art

[0002] An insurance policy is a formal contract between an insurance company and the insured, which details the insurance coverage, terms, liabilities and rights. It contains the insurance amount, insurance period, insurance type and related details, and is the legal evidence of the insurance relationship between the insured and the insurance company.

[0003] Currently, the management, review and quality inspection of insurance policies are still basically in the manual stage. This approach will lead to problems such as data entry errors, low processing efficiency, imperfect quality control, insufficient risk management, and inability to conduct comprehensive and compliant monitoring and tracking, which will affect the insured's insurance experience and will also be more likely to cause greater risk losses. Summary of the invention

[0004] The purpose of the present invention is to provide a comprehensive electronic quality inspection management platform for insurance policies, which can be achieved through the following technical solutions:

[0005] The embodiment of the present application provides a comprehensive electronic quality inspection management platform for insurance policies, including:

[0006] A database construction module is used to obtain insurance policy data, perform data preprocessing, and build an insurance policy database;

[0007] A data quality inspection module, which is used to create a quality inspection rule base and a verification rule base, and to obtain data analysis instructions and call the policy data in the policy database for data verification and data quality inspection;

[0008] The risk identification module is used to establish a risk identification model and conduct risk assessment on insurance policies based on the results of the quality inspection module;

[0009] Quality inspection and evaluation module, used to record and analyze quality indicators during data quality inspection and evaluate the quality of quality inspection;

[0010] Process monitoring module, used to monitor abnormalities and provide early warnings for quality inspection processes;

[0011] User management module, used to set access rights and operation rights according to user identity, perform data classification and encryption, and user access control;

[0012] Among them, the quality inspection module results include a first verification report, a second quality inspection report and a third repair report.

[0013] Preferably, the database construction module includes a data acquisition unit, a data processing unit and a database unit;

[0014] The data collection unit collects policy information from the data source and extracts original policy data;

[0015] The data processing unit pre-processes the original policy data to obtain policy data and generates a pre-processed data set based on the policy data;

[0016] The database unit establishes an association relationship between the pre-processed data set and the insurance policy information and stores them in an insurance policy database;

[0017] The insurance policy information includes contractual business information, preservation business information, claims business information and group insurance business information.

[0018] Preferably, the data processing unit includes a data checking subunit and a data classification subunit;

[0019] The data verification subunit is used to verify the integrity of the policy data and generate a verification data set, specifically:

[0020] Perform missing value detection on the policy data. If there are missing values, record the number and location of the missing values ​​and fill in the missing values.

[0021] Perform duplicate value detection on the policy data. If there are duplicate values, record the number of duplicate values ​​and delete the redundant duplicate values ​​after comparison;

[0022] Performing standardization test on the policy data, if there are any non-standardized values, converting the non-standardized values ​​into standardized values;

[0023] The data sets that have been verified for integrity are collected in the verification data set;

[0024] The data classification subunit classifies the data in the verification data set according to the data properties to generate text data, voice data and image data; if there is video data in the verification data set, the video data is converted into voice data and image data that are associated and corresponding to each other;

[0025] The data properties represent the characteristics and attributes of the data and are used to define the type, format, purpose and processing method of the data.

[0026] Preferably, the data quality inspection module includes a rule base construction unit, an instruction parsing unit, a data calling unit, a data verification unit, a data quality inspection unit and a data repair unit;

[0027] The rule base construction unit is used to construct a built-in rule base and a custom rule base and merge the two into a quality inspection rule base; it is also used to construct a verification rule base;

[0028] The instruction parsing unit is used to obtain the data analysis instruction and parse it into a data call request, a data verification request and a data quality inspection request;

[0029] The data calling unit is used to call the insurance policy data from the insurance policy database according to the data calling request;

[0030] The data verification unit performs data verification on the policy data based on the verification rule base and according to the data verification request and outputs a first verification report;

[0031] The data quality inspection unit performs data quality inspection on the policy data based on the quality inspection rule base and according to the data quality inspection request and outputs a second quality inspection report;

[0032] The data repair unit is used to perform data repair on the insurance policy data according to the first verification report and the second quality inspection report and output a third repair report;

[0033] The data call request includes a database identifier and a data identifier, and the policy data is called in the data call unit, specifically:

[0034] Selecting, according to the database identifier, an insurance policy database matching the database identifier;

[0035] The insurance policy data matching the data identifier is selected from the insurance policy database according to the data identifier.

[0036] Preferably, the data verification unit includes a data disassembly subunit, a data screening subunit, an abnormality judgment subunit and a policy verification subunit;

[0037] The data disassembly subunit is used to disassemble the policy data to obtain an abstract syntax tree of the policy data;

[0038] The data screening subunit screens the abstract syntax tree based on the verification rule base and obtains a screening result;

[0039] The abnormality judgment subunit is used to perform a first abnormality judgment based on the screening result to judge whether the insurance policy data corresponding to the abstract syntax tree belongs to normal insurance policy data; if it does, the normal insurance policy data is sent to the insurance policy verification subunit; if it does not, the abnormal insurance policy data is intercepted and sent to the data repair unit;

[0040] The insurance policy verification subunit is used to perform logic verification and format verification on the normal insurance policy data and output a first verification report;

[0041] Among them, in the insurance policy verification sub-unit, a second abnormality judgment is performed on the normal insurance policy data after verification. If the normal insurance policy data after verification is suspected abnormal insurance policy data, the suspected abnormal insurance policy data is saved in the suspected abnormal insurance policy database and waits for the next round of abnormality judgment in the abnormal judgment sub-unit.

[0042] Preferably, the data quality inspection unit includes a type judgment subunit, a voice-to-image conversion subunit and a text data quality inspection subunit;

[0043] The type judgment subunit is used to judge the type of the policy data. If the policy data is text data, it is sent to the text data quality inspection subunit; if the policy data is voice data and image data, it is sent to the voice-image conversion subunit;

[0044] The voice-to-image conversion subunit is used to convert the voice data and the image data into voice text and image text respectively and send the converted text to the text data quality inspection subunit;

[0045] The text data quality inspection subunit is used to collect text data, voice text and image text into a policy text set, perform quality inspection on the data in the policy text set and output a second quality inspection report.

[0046] Preferably, in the text data quality inspection subunit, quality inspection is performed on the data in the policy text set, including the following contents:

[0047] Selecting the text to be inspected from the policy text set;

[0048] Determine the text quality inspection points and scope;

[0049] Perform text quality inspection at the text quality inspection point and within the quality inspection range using a quality inspection rule library and obtain a second quality inspection report;

[0050] The text quality inspection points represent the places in the text to be inspected that do not conform to the quality inspection rules or contain errors;

[0051] Determine the text quality inspection points, specifically:

[0052] According to the preset quality inspection points and their corresponding quality inspection rules, a full-text search engine is used to select candidate quality inspection points that match the preset quality inspection points from the text to be inspected and mark them;

[0053] Using a preset neural network model, the text to be inspected is classified and the classification quality inspection points are marked;

[0054] Integrating the candidate quality check points and the classification quality check points to generate final text quality check points;

[0055] Determine the scope of quality inspection, specifically:

[0056] Parsing the text to be inspected to determine a text application scenario corresponding to the text to be inspected;

[0057] Selecting a keyword vector corresponding to the text application scenario;

[0058] Determine a first timestamp and a second timestamp corresponding to the keyword vector;

[0059] The quality inspection scope of the text to be quality inspected is determined according to the first timestamp and the second timestamp.

[0060] Preferably, the quality inspection and evaluation module includes a quality report generation unit and a visualization unit;

[0061] The quality report generation unit is used to record and analyze the quality indicators during data quality inspection and automatically generate a quality report based on the quality inspection results;

[0062] The visualization unit is used to perform visual analysis on the insurance policy data undergoing quality inspection based on the quality report and quality indicators.

[0063] Preferably, the process monitoring module includes a status monitoring unit and an abnormality early warning unit;

[0064] The status monitoring unit is used to monitor the quality inspection process of the insurance policy in real time, including the progress and status of each link in the quality inspection process;

[0065] The abnormal warning unit is used to automatically identify and warn abnormal behaviors or illegal behaviors that occur in the quality inspection process.

[0066] Preferably, the user management module includes a permission setting unit, a data encryption unit, an access control unit and an access recording unit;

[0067] The permission setting unit sets access rights and operation rights according to user identities;

[0068] The data encryption unit classifies the policy data according to the sensitivity of the policy data and encrypts the sensitive policy data therein;

[0069] The access control unit uses role-based access control technology to authenticate the identity and authority of the data accessor, and if the authentication is successful, the data is decrypted and the sensitive policy data is obtained;

[0070] The access recording unit is used to record the access and operation of the data accessor to the policy data;

[0071] Among them, the user identities include insurance agents, claims specialists, policyholders, beneficiaries and platform administrators.

[0072] The beneficial effects of the present invention are:

[0073] (1) The present invention obtains policy data from a data source through a database construction module and establishes a policy database after data preprocessing, thereby ensuring the integrity and accuracy of the collected data; performs data verification and data quality inspection on the preprocessed data through a data quality inspection module to ensure the accuracy, integrity and compliance of the policy information in the policy data, and at the same time, for scenarios where a large number of policies need to be processed quickly, it can efficiently complete the quality inspection task and improve the processing speed; establishes a risk identification model through a risk identification module and conducts risk assessment on the policy based on the quality inspection results to ensure the comprehensiveness of the policy information processing process, and also provides a basis for renewal decisions by identifying potential risk points; records and analyzes the quality indicators during data quality inspection and evaluates the quality of quality inspection through a quality inspection evaluation module to improve the accuracy of policy quality inspection; performs abnormal monitoring and early warning prompts on the quality inspection process through a process monitoring module to reduce potential risks and improve the security of the quality inspection process; sets access rights and operation rights according to user identities through a user management module, performs data classification and encryption and user access control, identifies user identities and obtains policy data for users based on the identification results, thereby ensuring the security of policy data, policy processes and user information.

[0074] (2) The present invention improves the efficiency and accuracy of insurance policy processing and the security of insurance policy data in an overall automated and intelligent manner, reduces potential risks, and provides strong support for the safe and stable development of the insurance industry.

[0075] (3) The present invention conducts comprehensive quality inspection and integrated management of the insurance policy processing flow to ensure the accuracy, completeness and compliance of the insurance policy information, and realizes comprehensive automated detection and review of the insurance policy by integrating advanced automatic speech recognition technology, optical character recognition technology and natural language processing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] For better understanding and implementation, the technical solution of the present application is described in detail below with reference to the accompanying drawings.

[0077] Figure 1 A schematic diagram of the structure of a comprehensive electronic quality inspection management platform for insurance policies provided in an embodiment of the present application;

[0078] Figure 2 A flowchart of the steps of a comprehensive electronic quality inspection management method for insurance policies provided in an embodiment of the present application. DETAILED DESCRIPTION

[0079] In order to further explain the technical means and effects taken by the present invention to achieve the predetermined invention purpose, exemplary embodiments will be described in detail here, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are only examples of methods and systems consistent with some aspects of the present application as detailed in the attached claims.

[0080] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to any or all possible combinations of one or more associated listed items.

[0081] The specific implementation methods, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0082] Example 1

[0083] See also Figure 1 The embodiment of the present application provides a comprehensive electronic quality inspection management platform for insurance policies, including:

[0084] A database construction module is used to obtain insurance policy data, perform data preprocessing, and build an insurance policy database;

[0085] A data quality inspection module, which is used to create a quality inspection rule base and a verification rule base, and to obtain data analysis instructions and call the policy data in the policy database for data verification and data quality inspection;

[0086] The risk identification module is used to establish a risk identification model and conduct risk assessment on insurance policies based on the results of the quality inspection module;

[0087] Quality inspection and evaluation module, used to record and analyze quality indicators during data quality inspection and evaluate the quality of quality inspection;

[0088] Process monitoring module, used to monitor abnormalities and provide early warnings for quality inspection processes;

[0089] User management module, used to set access rights and operation rights according to user identity, perform data classification and encryption, and user access control;

[0090] Among them, the quality inspection module results include a first verification report, a second quality inspection report and a third repair report.

[0091] Specifically, the present application obtains policy data from a data source through a database construction module and establishes a policy database after data preprocessing, thereby ensuring the integrity and accuracy of the collected data; performs data verification and data quality inspection on the preprocessed data through a data quality inspection module to ensure the accuracy, integrity and compliance of the policy information in the policy data, and at the same time, for scenarios where a large number of policies need to be processed quickly, it can efficiently complete the quality inspection task and improve the processing speed; establishes a risk identification model through a risk identification module and conducts risk assessment on the policy based on the quality inspection results to ensure the comprehensiveness of the policy information processing process, and also provides a basis for renewal decisions by identifying potential risk points; records and analyzes quality indicators during data quality inspection and evaluates quality inspection quality through a quality inspection evaluation module to improve the accuracy of policy quality inspection; performs abnormal monitoring and early warning prompts on the quality inspection process through a process monitoring module to reduce potential risks and improve the security of the quality inspection process; sets access rights and operation rights according to user identity through a user management module, performs data classification and encryption and user access control, identifies user identity and obtains policy data for the user based on the identification results, thereby ensuring the security of policy data, policy process and user information. This application uses automation and intelligent methods to improve the efficiency and accuracy of policy processing and the security of policy data, reduce potential risks and provide strong support for the safe and stable development of the insurance industry.

[0092] In an embodiment provided in the present application, the database construction module includes a data acquisition unit, a data processing unit and a database unit;

[0093] The data collection unit collects policy information from the data source and extracts original policy data;

[0094] The data processing unit pre-processes the original policy data to obtain policy data and generates a pre-processed data set based on the policy data;

[0095] The database unit establishes an association relationship between the pre-processed data set and the insurance policy information and stores them in an insurance policy database;

[0096] The insurance policy information includes contractual business information, preservation business information, claims business information and group insurance business information.

[0097] Specifically, in the database construction module of the present embodiment, the data collection unit is first used to collect policy information from the data source and extract original policy data from the policy information. The data processing unit is then used to preprocess the original policy data to generate a preprocessed data set. Finally, the database unit is used to establish an association relationship between the preprocessed data set and the policy information to associate the two together, and the associated data and information are stored in the policy database. The present embodiment realizes the collection and processing of policy information and original policy data through the above-mentioned units, thereby ensuring that the final policy data is more complete and accurate.

[0098] It can be understood that establishing an association relationship between the preprocessed data set and the insurance policy information means establishing an association relationship between the preprocessed insurance policy data and the original insurance policy information so that there can be a corresponding relationship between the two, that is: the original data can be queried when needed through the association relationship between the preprocessed data and the original data.

[0099] It should be noted that the above-mentioned original policy data refers to data that has not been preprocessed, while the subsequent policy data refers to preprocessed policy data. In order to distinguish the two types of data before and after preprocessing in this application, the original policy data is used to represent the data before processing, and the policy data is used to represent the data after processing. Similarly, the policy data mentioned subsequently refer to the policy data obtained after data processing, and all the data in the policy database are also processed policy data.

[0100] It should also be noted that the above-mentioned data source refers to the source of policy information and policy data, which may include but is not limited to the insurance platform, insurance company system, third-party data source (credit assessment agencies, medical institutions, traffic management agencies, etc.), among which the third-party data source is related to the type of insurance the user purchases. For example, if the customer's insurance type is traffic insurance, the third-party data source is the Traffic Management Bureau or institutions and departments related to traffic insurance claims. These institutions and departments will provide relevant accident data as reference data for insurance claims, so these data are also part of the policy information. The customer's insurance type can also be medical insurance or health insurance, etc., which will not be elaborated here.

[0101] It can be understood that the insurance policy information in this embodiment includes but is not limited to contract business information, preservation business information, claims business information and group insurance business information, specifically insurance business related information provided by the insurance company.

[0102] Among them, the contract business mainly provides basic information of data insurance policies and related insurance policy images, as well as review and operation functions. The basic information includes the following:

[0103] Insurance type information: insurance contract number, application form number, application form status, insurance amount, new policy premium, regular insurance, accumulated premium, insurance type code, insurance type name, insurance type status, insurance type category, primary and supplementary insurance mark, long and short insurance mark, insurance term, payment year, payment method, payment interval, name of the insured, gender of the insured, insurance age of the insured, name of the insured, gender of the insured, insurance age of the insured.

[0104] Salesman and channel information: salesman code, salesman name, main sales channel, sales sub-channel, third-level channel_source of insurance, channel salesman name, third-level management agency name, fourth-level management agency name, sales BU, partner name, partner agency name, channel merchant name, channel merchant agency name, whether self-operated, whether in Shenzhen Hefei District, whether domestic non-resident, assessment agency, sales agency code, sales agency name, business department name, prepayment date, underwriting date, underwriting time, effective date, customer receipt date, receipt return date, pre-insurance withdrawal date, cooling-off period cancellation date, cancellation date, cancellation amount, and follow-up success date.

[0105] The preservation business mainly provides basic information of data insurance policies and related insurance application images, as well as review and operation functions. The basic information includes: insurance contract number, insurance application number, insurance application status, insurance amount, policyholder name, policyholder gender, policyholder insurance age, insured name, insured gender, insured age, relationship between policyholder and insured, and related preservation project information that has been applied for; the image content needs to include a preservation application, a copy of the insured's identity, a power of attorney application, etc.

[0106] In one embodiment provided in the present application, the data processing unit includes a data checking subunit and a data classification subunit;

[0107] The data verification subunit is used to verify the integrity of the policy data and generate a verification data set, specifically:

[0108] Perform missing value detection on the policy data. If there are missing values, record the number and location of the missing values ​​and fill in the missing values.

[0109] Duplicate value detection is performed on the insurance policy data. If duplicate values ​​exist, the number of duplicate values ​​is recorded and the redundant duplicate values ​​are deleted after comparison; only the most complete one among multiple duplicate values ​​is retained.

[0110] Performing standardization test on the policy data, if there are any non-standardized values, converting the non-standardized values ​​into standardized values;

[0111] The data sets that have been verified for integrity are collected in the verification data set;

[0112] The data classification subunit classifies the data in the verification data set according to the data properties to generate text data, voice data and image data; if there is video data in the verification data set, the video data is converted into voice data and image data that are associated and corresponding to each other;

[0113] Among them, the above-mentioned data properties represent the characteristics and attributes of the data, and are used to define the type, format, purpose and processing method of the data.

[0114] Specifically, this embodiment performs a preliminary check on the integrity of the insurance policy data through the data verification sub-unit, and collects the verified data into a verification data set to provide data support for subsequent data classification and quality inspection. In addition, the collected data can be preliminarily improved through data verification, and the data from different data sources can be processed and aggregated according to unified standards, which is convenient for subsequent data analysis and decision support and can ensure the integrity of the data.

[0115] It can be understood that in standardized testing, it is necessary to detect whether there are data of different formats or standards in the data. Non-standardized values ​​refer to data that differ from standard values. The definition of standard values ​​is set according to actual needs. This embodiment only uses standardized testing as a means and does not limit the specific content of the standard values.

[0116] In an embodiment provided in the present application, the data quality inspection module includes a rule base construction unit, an instruction parsing unit, a data calling unit, a data verification unit, a data quality inspection unit and a data repair unit;

[0117] The rule base construction unit is used to construct a built-in rule base and a custom rule base and merge the two into a quality inspection rule base; it is also used to construct a verification rule base;

[0118] The instruction parsing unit is used to obtain the data analysis instruction and parse it into a data call request, a data verification request and a data quality inspection request;

[0119] The data calling unit is used to call the insurance policy data from the insurance policy database according to the data calling request;

[0120] The data verification unit performs data verification on the policy data based on the verification rule base and according to the data verification request and outputs a first verification report;

[0121] The data quality inspection unit performs data quality inspection on the policy data based on the quality inspection rule base and according to the data quality inspection request and outputs a second quality inspection report;

[0122] The data repair unit is used to perform data repair on the insurance policy data according to the first verification report and the second quality inspection report and output a third repair report;

[0123] The data call request includes a database identifier and a data identifier, and the policy data is called in the data call unit, specifically:

[0124] Selecting, according to the database identifier, an insurance policy database matching the database identifier;

[0125] The insurance policy data matching the data identifier is selected from the insurance policy database according to the data identifier.

[0126] Specifically, this embodiment first constructs a built-in rule base and a custom rule base and merges the two into a quality inspection rule base, and also constructs a verification rule base for data verification; then, multiple data requests are obtained by parsing data analysis instructions, and subsequent different data operations are performed according to different data requests, such as calling policy data from the policy database through a data call request, performing data verification on the policy data through a data verification request and based on the verification rule base, performing data quality inspection on the policy data through a data quality inspection request and based on the quality inspection rule base, and finally repairing the policy data according to the report obtained after verification and quality inspection, thereby ensuring the accuracy and integrity of the policy data while improving the accuracy of data verification and quality inspection.

[0127] It is understandable that the construction of the above-mentioned quality inspection rule base and verification rule base are based on the platform's insurance policy database and the rule base set according to actual conditions. Specifically, the built-in rule base is a set of built-in compliance inspection rules provided by the platform, which are formulated based on industry standards and internal requirements of insurance companies; and the custom rule base means: allowing users to customize compliance inspection rules according to specific business needs to meet personalized quality inspection needs. It can be seen that this embodiment also makes the platform flexible and scalable through the above-mentioned quality inspection rule base, among which flexibility is manifested in: allowing users to flexibly adjust various parameters and settings of the platform according to actual needs; scalability is manifested in: allowing functional expansion and customized development according to the business development needs of insurance companies.

[0128] In an embodiment provided in the present application, the data verification unit includes a data disassembly subunit, a data screening subunit, an abnormality judgment subunit and a policy verification subunit;

[0129] The data disassembly subunit is used to disassemble the policy data to obtain an abstract syntax tree of the policy data;

[0130] The data screening subunit screens the abstract syntax tree based on the verification rule base and obtains a screening result;

[0131] The abnormality judgment subunit is used to perform a first abnormality judgment based on the screening result to judge whether the insurance policy data corresponding to the abstract syntax tree belongs to normal insurance policy data; if it does, the normal insurance policy data is sent to the insurance policy verification subunit; if it does not, the abnormal insurance policy data is intercepted and sent to the data repair unit;

[0132] The insurance policy verification subunit is used to perform logic verification and format verification on the normal insurance policy data and output a first verification report;

[0133] Among them, in the insurance policy verification sub-unit, a second abnormality judgment is performed on the normal insurance policy data after verification. If the normal insurance policy data after verification is suspected abnormal insurance policy data, the suspected abnormal insurance policy data is saved in the suspected abnormal insurance policy database and waits for the next round of abnormality judgment in the abnormal judgment sub-unit.

[0134] In an embodiment provided by the present application, when performing the next round of abnormality judgment, if the suspected abnormal policy data meets the business rules of normal policy data, it is treated as normal policy data for subsequent processing; if the suspected abnormal policy data does not meet the business rules of normal policy data, it is treated as abnormal policy data for subsequent processing;

[0135] Among them, the business rules of normal insurance policy data refer to the rules that the insurance policy data has when it is normally executed in the insurance policy business process.

[0136] Specifically, the accuracy of the policy data is ensured by further judging the suspected abnormal policy data. At the same time, the specific content of the business rules for normal policy data is determined according to the actual situation. Its role is to ensure that the policy data can be used normally in the business process for insurance business such as insurance, claims settlement and preservation.

[0137] It is understandable that data verification is to verify the integrity and correctness of the policy data, including but not limited to: data integrity check: the platform will automatically check the integrity of the information in the policy to ensure that there are no missing or missing key fields; data accuracy check: logical verification and format verification of the data in the policy to identify and correct potential errors or inconsistencies. This embodiment can ensure the integrity and accuracy of the policy information through the policy verification and reduce the risk of subsequent claims.

[0138] In one embodiment provided in the present application, the data quality inspection unit includes a type judgment subunit, a voice-to-image conversion subunit, and a text data quality inspection subunit;

[0139] The type judgment subunit is used to judge the type of the policy data. If the policy data is text data, it is sent to the text data quality inspection subunit; if the policy data is voice data and image data, it is sent to the voice-image conversion subunit;

[0140] The voice-to-image conversion subunit is used to convert the voice data and the image data into voice text and image text respectively and send the converted text to the text data quality inspection subunit;

[0141] The text data quality inspection subunit is used to collect text data, voice text and image text into a policy text set, perform quality inspection on the data in the policy text set and output a second quality inspection report.

[0142] Specifically, in the data quality inspection unit, first determine which type the policy data belongs to. If it is text data, directly send it to the text data quality inspection subunit for quality inspection; if it is voice data or image data, convert the above two types of data into text data and send the converted data to the text data quality inspection subunit for quality inspection. In summary, in this embodiment, the type judgment subunit is used to judge and divide the type of policy data and perform different operations, thereby ensuring that different types of data can receive different operations, and ensuring the comprehensiveness and accuracy of data quality inspection.

[0143] In an embodiment provided in the present application, in the voice-to-image conversion subunit, converting the voice data into voice text includes:

[0144] 1) Train the speech recognition model, specifically:

[0145] Collect multilingual speech data including Chinese language and foreign language and perform speech preprocessing; among which, Chinese language includes Mandarin language and dialect language;

[0146] Extracting speech feature parameters corresponding to multilingual speech data and generating a speech feature set;

[0147] Extracting a preset proportion of speech feature parameters from the speech feature set, and using them to construct an initial speech recognition model;

[0148] Training the initial speech recognition model through an iterative algorithm to obtain optimal model parameters;

[0149] The speech recognition model corresponding to the optimal model parameters obtained through training is used as the final speech recognition model;

[0150] 2) Inputting the speech data into the speech recognition model to obtain speech text, specifically:

[0151] Converting speech data into a sound wave graph, and obtaining sound wave-speech feature parameters corresponding to the sound wave graph;

[0152] Inputting the sound wave-speech feature parameters into the speech recognition model to perform language family recognition to obtain the language family to which the speech data belongs;

[0153] Speech recognition is performed according to the language family to which it belongs, and speech feature parameters corresponding to the language family are obtained and processed to obtain the speech text corresponding to the speech data.

[0154] Specifically, in this embodiment, the multilingual speech data refers to speech data of multiple languages ​​including the Chinese language and foreign language, and the above-mentioned Chinese language includes Mandarin language and dialect language, wherein the foreign language refers to language systems other than Chinese, such as English, French, German, etc.; the multilingual system is set in this embodiment in order to have richer training parameters in model training, so as to ensure that the subsequently obtained speech recognition model is more accurate and comprehensive; and this embodiment constructs an initial speech recognition model by extracting a preset proportion of speech feature parameters, which can make the model more professional and characteristic, such as: setting the proportion of speech feature parameters of the Chinese language to be higher than that of the foreign language, the obtained initial speech recognition model will give priority to recognizing speech data of the Chinese language, and will be more inclined to the recognition of Mandarin and dialects, otherwise it will not be repeated.

[0155] In an embodiment provided in the present application, in the speech-to-image conversion subunit, converting image data into image text includes the following contents:

[0156] Acquire image data and perform feature extraction to detect the location of text information in the image data;

[0157] Using OCR technology to perform text recognition at the text information location to obtain initial text information;

[0158] Performing content correction on the initial text information in combination with scene information and text semantics of the image data;

[0159] The corrected text information is formatted to obtain the final image text.

[0160] Specifically, OCR (Optical Character Recognition) technology refers to optical character recognition technology, which can be used to recognize text content in an image and convert it into editable text.

[0161] In an embodiment provided by the present application, in the text data quality inspection subunit, quality inspection is performed on the data in the insurance policy text set, including the following contents:

[0162] Selecting the text to be inspected from the policy text set;

[0163] Determine the text quality inspection points and scope;

[0164] Perform text quality inspection at the text quality inspection point and within the quality inspection range using a quality inspection rule library and obtain a second quality inspection report;

[0165] The text quality inspection points represent the places in the text to be inspected that do not conform to the quality inspection rules or contain errors;

[0166] Determine the text quality inspection points, specifically:

[0167] According to the preset quality inspection points and their corresponding quality inspection rules, using a full-text search engine, selecting candidate quality inspection points that match the preset quality inspection points from the text to be inspected and marking them;

[0168] Using a preset neural network model, the text to be inspected is classified and the classification quality inspection points are marked;

[0169] Integrating the candidate quality check points and the classification quality check points to generate final text quality check points;

[0170] Determine the scope of quality inspection, specifically:

[0171] Parsing the text to be inspected to determine a text application scenario corresponding to the text to be inspected;

[0172] Selecting a keyword vector corresponding to the text application scenario;

[0173] Determine a first timestamp and a second timestamp corresponding to the keyword vector;

[0174] The quality inspection scope of the text to be quality inspected is determined according to the first timestamp and the second timestamp.

[0175] Specifically, regarding the determination of text quality inspection points, this embodiment determines them by integrating candidate quality inspection points and classified quality inspection points, thereby improving the accuracy of quality inspection point confirmation; text quality inspection points represent the places in the text to be inspected that are non-compliant or illegal compared with the quality inspection rules, which can also be understood as places where errors exist; and the quality inspection scope is the scope of text quality inspection determined by keyword vectors based on the text application scenario, which can well represent the scope of quality inspection required for the text to be inspected. Among them, the text application scenario refers to the specific environment or field where the text data is applied and used.

[0176] In one embodiment provided in the present application, the quality inspection and evaluation module includes a quality report generating unit and a visualization unit;

[0177] The quality report generation unit is used to record and analyze the quality indicators during data quality inspection, and automatically generate a quality report based on the quality inspection results; the quality report includes information such as error rate and abnormal distribution;

[0178] The visualization unit is used to perform visual analysis on the insurance policy data undergoing quality inspection based on the quality report and quality indicators.

[0179] Specifically, in the above two units, generating quality reports means: automatically generating detailed quality reports based on quality inspection results, including error rates, abnormal distribution and other information; data analysis and visualization means: in-depth analysis and visual display of quality inspection data to help management understand the quality status of policy processing, identify potential problems and make improvements.

[0180] In an embodiment provided in the present application, the process monitoring module includes a state monitoring unit and an abnormality warning unit;

[0181] The status monitoring unit is used to monitor the quality inspection process of the insurance policy in real time, including the progress and status of each link in the quality inspection process;

[0182] The abnormal warning unit is used to automatically identify and warn abnormal behaviors or illegal behaviors that occur in the quality inspection process.

[0183] Specifically, real-time process monitoring: the platform can monitor the policy processing process in real time, including the progress and status of each link; abnormality identification and reminder: for abnormalities or violations that occur in the process, the platform can automatically identify and remind relevant personnel to handle them.

[0184] In an embodiment provided in the present application, the user management module includes a permission setting unit, a data encryption unit, an access control unit and an access recording unit;

[0185] The permission setting unit sets access rights and operation rights according to user identities;

[0186] The data encryption unit classifies the policy data according to the sensitivity of the policy data and encrypts the sensitive policy data therein;

[0187] The access control unit uses role-based access control technology to authenticate the identity and authority of the data accessor, and if the authentication is successful, the data is decrypted and the sensitive policy data is obtained;

[0188] The access recording unit is used to record the access and operation of the data accessor to the policy data;

[0189] Among them, the user identities include insurance agents, claims specialists, policyholders, beneficiaries and platform administrators.

[0190] Specifically, in this embodiment, first, specific access rights and operation rights (such as viewing, editing, and deleting insurance policies) are set for each identity role according to the user identity; then, the insurance policy data is classified according to the sensitivity of the insurance policy data, and it is divided into sensitive insurance policy data (such as personal information, insurance policy amount, etc.) and non-sensitive insurance policy data, and the sensitive insurance policy data is encrypted; then, the role-based access control technology (RBAC, Role-Based Access Control) is used to authenticate the identity and authority of the data accessor, and if the verification is passed, the data is decrypted and the sensitive insurance policy data is obtained; finally, the access and operation of the data accessor to the insurance policy data is recorded in real time for easy auditing and monitoring; wherein, the user identity includes insurance agent, claim specialist, policyholder, beneficiary, and platform administrator. Through the above steps and contents, this embodiment can ensure that the insurance policy data in the insurance business is effectively protected in terms of authority management and data encryption, and maintain the security and integrity of the insurance policy data. It should be noted that this embodiment also implements encryption during data storage and transmission to ensure the security of data throughout the life cycle (insurance policy business process).

[0191] In another embodiment provided by the present application, the above-mentioned comprehensive electronic quality inspection management platform for insurance policies is applicable to various insurance companies and insurance intermediary institutions, and plays an important role in the following scenarios in particular:

[0192] New policy underwriting: Conduct quality inspection before policy underwriting to ensure the integrity and accuracy of policy information and reduce subsequent claims risks.

[0193] Policy renewal: Conduct quality inspection on renewal policies, identify potential risk points, and provide a basis for renewal decisions.

[0194] Batch policy processing: For scenarios where a large number of policies need to be processed quickly, the platform can efficiently complete quality inspection tasks and improve processing speed.

[0195] In short, the above-mentioned comprehensive electronic quality inspection management platform for insurance policies is an intelligent platform system designed specifically for the insurance industry, aiming to improve the efficiency of insurance policy processing, ensure the quality of insurance policies and reduce potential risks. The functions of the above-mentioned quality inspection management platform mainly focus on comprehensive quality inspection and management of the insurance policy processing process to ensure the accuracy, completeness and compliance of insurance policy information. The platform integrates advanced automatic speech recognition technology, optical character recognition technology and natural language processing technology to achieve comprehensive automated detection and review of insurance policies.

[0196] In summary, the above-mentioned comprehensive electronic quality inspection management platform for insurance policies improves the efficiency and accuracy of insurance policy processing through automation and intelligent methods, reduces potential risks, and provides strong support for the stable development of the insurance industry.

[0197] Example 2

[0198] See also Figure 2 The embodiment of the present application provides a comprehensive electronic quality inspection management method for insurance policies, which is applied to the comprehensive electronic quality inspection management platform for insurance policies as described above, and specifically includes the following steps:

[0199] Obtain insurance policy data, perform data preprocessing, and build an insurance policy database;

[0200] Creating a quality inspection rule base and a verification rule base, and obtaining data analysis instructions and calling the policy data in the policy database for data verification and data quality inspection;

[0201] Establish a risk identification model and conduct risk assessment on insurance policies based on the results of the quality inspection module;

[0202] Record and analyze quality indicators during data quality inspection and evaluate the quality of quality inspection;

[0203] Monitor abnormalities and provide early warnings for quality inspection processes;

[0204] Set access rights and operation rights according to user identities, perform data classification and encryption, and control user access;

[0205] Among them, the quality inspection module results include a first verification report, a second quality inspection report and a third repair report.

[0206] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0207] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0208] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0209] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A comprehensive electronic quality inspection management platform for insurance policies, characterized by: include: A database construction module is used to obtain insurance policy data, perform data preprocessing, and build an insurance policy database; A data quality inspection module, which is used to create a quality inspection rule base and a verification rule base, and to obtain data analysis instructions and call the policy data in the policy database for data verification and data quality inspection; The risk identification module is used to establish a risk identification model and conduct risk assessment on insurance policies based on the results of the quality inspection module; Quality inspection and evaluation module, used to record and analyze quality indicators during data quality inspection and evaluate the quality of quality inspection; Process monitoring module, used to monitor abnormalities and provide early warnings for quality inspection processes; User management module, used to set access rights and operation rights according to user identity, perform data classification and encryption, and user access control; Among them, the quality inspection module results include a first verification report, a second quality inspection report and a third repair report.

2. The comprehensive electronic quality inspection management platform for insurance policies according to claim 1 is characterized by: The database construction module includes a data acquisition unit, a data processing unit and a database unit; The data collection unit collects policy information from the data source and extracts original policy data; The data processing unit pre-processes the original policy data to obtain policy data and generates a pre-processed data set based on the policy data; The database unit establishes an association relationship between the pre-processed data set and the insurance policy information and stores them in an insurance policy database; The insurance policy information includes contractual business information, preservation business information, claims business information and group insurance business information.

3. The comprehensive electronic quality inspection management platform for insurance policies according to claim 2 is characterized by: The data processing unit includes a data checking subunit and a data classification subunit; The data verification subunit is used to verify the integrity of the policy data and generate a verification data set, specifically: Perform missing value detection on the policy data. If there are missing values, record the number and location of the missing values ​​and fill in the missing values. Perform duplicate value detection on the policy data. If there are duplicate values, record the number of duplicate values ​​and delete the redundant duplicate values ​​after comparison; Performing standardization test on the policy data, if there are any non-standardized values, converting the non-standardized values ​​into standardized values; The data sets that have been verified for integrity are collected in the verification data set; The data classification subunit classifies the data in the verification data set according to the data properties to generate text data, voice data and image data; if there is video data in the verification data set, the video data is converted into voice data and image data that are associated and corresponding to each other; The data properties represent the characteristics and attributes of the data and are used to define the type, format, purpose and processing method of the data.

4. The comprehensive electronic quality inspection management platform for insurance policies according to claim 1 is characterized by: The data quality inspection module includes a rule base construction unit, an instruction parsing unit, a data calling unit, a data verification unit, a data quality inspection unit and a data repair unit; The rule base construction unit is used to construct a built-in rule base and a custom rule base and merge the two into a quality inspection rule base; it is also used to construct a verification rule base; The instruction parsing unit is used to obtain the data analysis instruction and parse it into a data call request, a data verification request and a data quality inspection request; The data calling unit is used to call the insurance policy data from the insurance policy database according to the data calling request; The data verification unit performs data verification on the policy data based on the verification rule base and according to the data verification request and outputs a first verification report; The data quality inspection unit performs data quality inspection on the policy data based on the quality inspection rule base and according to the data quality inspection request and outputs a second quality inspection report; The data repair unit is used to perform data repair on the insurance policy data according to the first verification report and the second quality inspection report and output a third repair report; The data call request includes a database identifier and a data identifier, and the policy data is called in the data call unit, specifically: Selecting, according to the database identifier, an insurance policy database matching the database identifier; The insurance policy data matching the data identifier is selected from the insurance policy database according to the data identifier.

5. The comprehensive electronic quality inspection management platform for insurance policies according to claim 4 is characterized by: The data verification unit includes a data disassembly subunit, a data screening subunit, an abnormality judgment subunit and a policy verification subunit; The data disassembly subunit is used to disassemble the policy data to obtain an abstract syntax tree of the policy data; The data screening subunit screens the abstract syntax tree based on the verification rule base and obtains a screening result; The abnormality judgment subunit is used to perform a first abnormality judgment according to the screening result, and judge whether the insurance policy data corresponding to the abstract syntax tree belongs to normal insurance policy data; If yes, the normal policy data is sent to the policy verification subunit; If not, intercepting the abnormal policy data and sending it to the data repair unit; The insurance policy verification subunit is used to perform logic verification and format verification on the normal insurance policy data and output a first verification report; Among them, in the insurance policy verification sub-unit, a second abnormality judgment is performed on the normal insurance policy data after verification. If the normal insurance policy data after verification is suspected abnormal insurance policy data, the suspected abnormal insurance policy data is saved in the suspected abnormal insurance policy database and waits for the next round of abnormality judgment in the abnormal judgment sub-unit.

6. The comprehensive electronic quality inspection management platform for insurance policies according to claim 4 is characterized by: The data quality inspection unit includes a type judgment subunit, a voice-to-image conversion subunit and a text data quality inspection subunit; The type judgment subunit is used to judge the type of the policy data. If the policy data is text data, it is sent to the text data quality inspection subunit; if the policy data is voice data and image data, it is sent to the voice-image conversion subunit; The voice-to-image conversion subunit is used to convert the voice data and the image data into voice text and image text respectively and send the converted text to the text data quality inspection subunit; The text data quality inspection subunit is used to collect text data, voice text and image text into a policy text set, perform quality inspection on the data in the policy text set and output a second quality inspection report.

7. The comprehensive electronic quality inspection management platform for insurance policies according to claim 6 is characterized by: In the text data quality inspection subunit, the data in the policy text set is quality inspected, including the following contents: Selecting the text to be inspected from the policy text set; Determine the text quality inspection points and scope; Perform text quality inspection at the text quality inspection point and within the quality inspection range using a quality inspection rule library and obtain a second quality inspection report; The text quality inspection points represent the places in the text to be inspected that do not conform to the quality inspection rules or contain errors; Determine the text quality inspection points, specifically: According to the preset quality inspection points and their corresponding quality inspection rules, a full-text search engine is used to select candidate quality inspection points that match the preset quality inspection points from the text to be inspected and mark them; Using a preset neural network model, the text to be inspected is classified and the classification quality inspection points are marked; Integrating the candidate quality check points and the classification quality check points to generate final text quality check points; Determine the scope of quality inspection, specifically: Parsing the text to be inspected to determine a text application scenario corresponding to the text to be inspected; Selecting a keyword vector corresponding to the text application scenario; Determine a first timestamp and a second timestamp corresponding to the keyword vector; The quality inspection scope of the text to be quality inspected is determined according to the first timestamp and the second timestamp.

8. The comprehensive electronic quality inspection management platform for insurance policies according to claim 1 is characterized by: The quality inspection and evaluation module includes a quality report generation unit and a visualization unit; The quality report generation unit is used to record and analyze the quality indicators during data quality inspection and automatically generate a quality report based on the quality inspection results; The visualization unit is used to perform visual analysis on the insurance policy data undergoing quality inspection based on the quality report and quality indicators.

9. The comprehensive electronic quality inspection management platform for insurance policies according to claim 1 is characterized by: The process monitoring module includes a status monitoring unit and an abnormality early warning unit; The status monitoring unit is used to monitor the quality inspection process of the insurance policy in real time, including the progress and status of each link in the quality inspection process; The abnormal warning unit is used to automatically identify and warn abnormal behaviors or illegal behaviors that occur in the quality inspection process.

10. The comprehensive electronic quality inspection management platform for insurance policies according to claim 1 is characterized by: The user management module includes a permission setting unit, a data encryption unit, an access control unit and an access recording unit; The permission setting unit sets access rights and operation rights according to user identities; The data encryption unit classifies the policy data according to the sensitivity of the policy data and encrypts the sensitive policy data therein; The access control unit uses role-based access control technology to authenticate the identity and authority of the data accessor, and if the authentication is successful, the data is decrypted and the sensitive policy data is obtained; The access recording unit is used to record the access and operation of the data accessor to the policy data; Among them, the user identities include insurance agents, claims specialists, policyholders, beneficiaries and platform administrators.

Images