Method and system for screening and positioning insured expansion personnel based on multi-source data fusion
Through multi-source data fusion and multi-layer data verification technology, the problem of single data sources and large manual sorting workload in insurance expansion has been solved, and the precise positioning and accurate allocation of insurance expansion personnel has been achieved, and the expansion work has been improved. efficiency and accuracy.
Patent Information
- Application Number
- CN202410974835.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-19
AI Technical Summary
The existing insurance coverage expansion methods have problems such as single data source, large manual sorting workload, poor screening accuracy and inaccurate allocation of people participating in insurance coverage expansion, resulting in low efficiency and low accuracy of expansion.
By using a multi-source data fusion method, by obtaining multi-channel data, extracting interrupted payers, analyzing potential expansion databases, extracting and word segmentation address information, conducting multi-layer data verification and matching, and positioning the medical insurance institutions to which the insurance expansion personnel belong. Realize accurate data allocation.
It improves the accuracy of screening of insurance and expanded personnel and the accuracy of data allocation, reduces the workload of business personnel, and ensures the authenticity and accuracy of expanded data.
Smart Images

Figure CN118981509A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of medical insurance data processing, and specifically relates to a method and system for screening and locating insured persons based on multi-source data fusion. Background Art
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] In order to improve the basic medical security system and expand the coverage of medical insurance, the work of expanding the coverage of insurance has been gradually put on the agenda. The purpose of expanding the coverage of insurance is to expand the scale and scope of insurance for those who are not insured. At present, in the process of expanding the coverage of insurance, some potential insured persons can be identified through the background query of medical insurance data by business personnel.
[0004] According to the inventors' understanding, the existing insurance coverage expansion methods have problems such as single data source, large workload of manual sorting, and poor accuracy in locating the expanded population when processing data. These problems will cause difficulties for grassroots personnel in carrying out this work, affecting the progress of the expansion work; the specific problems are as follows:
[0005] (1) Single source of data screening: In the process of expanding the coverage of insurance, data comparisons are required from multiple departments. However, the traditional data screening method has a single source of data and only relies on internal medical insurance data for screening and analysis. It is unable to identify and delete invalid expansion data such as death and normal insurance coverage in other places.
[0006] (2) The workload of manual sorting is large: Traditional expanded screening work relies heavily on manual processing. The amount of data that needs to be analyzed for universal insurance is huge, and data integration and data comparison require a lot of manual conversion and sorting work, resulting in low efficiency and low accuracy.
[0007] (3) Poor screening accuracy for the insured population: Due to incomplete data aggregation and lack of multi-layer data screening, the target group is too large, affecting the screening accuracy of the insured population.
[0008] (4) Inaccurate allocation of insured persons: Data sources are diverse, and the quality of data varies from region to region and department to department. There are problems such as inaccurate and incomplete data and inconsistent formats, which affect the accuracy of data allocation and make it impossible to accurately allocate selected insured persons to their respective institutions. Summary of the invention
[0009] To solve the above problems, the present invention proposes a screening and positioning method and system for insurance coverage expansion personnel based on multi-source data fusion. Multi-source data fusion is performed on the collected multi-channel data to ensure the comprehensiveness of data screening, comprehensively analyze the potential for personnel expansion, effectively improve the accuracy of expansion data screening and data allocation, and reduce the workload of business personnel; through the analysis and matching of personnel address information, structured processing of address information is achieved, and the accuracy of positioning of insurance coverage expansion personnel is improved.
[0010] According to some embodiments, the first solution of the present invention provides a method for screening and locating insurance coverage expansion personnel based on multi-source data fusion, using the following technical solutions:
[0011] A screening and positioning method for insurance coverage expansion personnel based on multi-source data fusion, comprising:
[0012] Obtain multi-channel data information of personnel to be processed;
[0013] Based on the multi-channel data information obtained, the personnel who have stopped paying are extracted from the personnel to be processed;
[0014] Analyze the data information of the proposed personnel who have stopped paying premiums to obtain a potential expansion database;
[0015] Analyze the potential expansion database to obtain the uninsured data that can be expanded;
[0016] Extract the address information fields, address segmentation and coding classification of the obtained expandable uninsured data to obtain the standard address information of the expandable uninsured persons;
[0017] The obtained standard address information is compared and matched with the medical insurance agency information in the local database to obtain the medical insurance agency to which the uninsured persons who can be expanded to cover the insurance coverage belong, thereby completing the positioning of the insured persons who can be expanded to cover the insurance coverage.
[0018] As a further technical limitation, based on the multi-channel data information obtained, the persons who have interrupted payment are extracted from the persons to be processed; based on the data characteristics of the extracted persons who have interrupted payment, target data including at least the insured status, age, benefit enjoyment, payment period, insurance interruption and payment interruption are obtained; the obtained target data are classified, and the multi-channel data information obtained is screened regularly according to the target data and its weight to obtain a potential expansion database.
[0019] As a further technical limitation, the obtained potential expansion database is subject to multi-layer data verification, and the non-expandable data in the potential expansion database are eliminated in turn to obtain expandable uninsured data. As a further technical limitation, the address information fields of the expandable uninsured data are extracted through address information aggregation, data cleaning and rule definition, and the extracted address information fields are subjected to address segmentation and matching calculation to obtain the geographical elements of the address information fields; the obtained geographical elements of the address information fields are coded and converted to obtain the standard address information of the expandable uninsured persons.
[0020] As a further technical limitation, the obtained standard address information is matched with the medical insurance institution information maintained in the local database, and the medical insurance institution information that matches the standard address information is screened out, that is, the medical insurance institutions to which the uninsured persons who can be expanded belong are obtained, and the medical insurance institutions to which the uninsured persons who can be expanded belong are located, thereby achieving the positioning of the insured persons who can be expanded.
[0021] As a further technical limitation, the multi-channel data information obtained includes at least local medical insurance data, external medical insurance data and shared data.
[0022] According to some embodiments, the second solution of the present invention provides a screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion, which adopts the following technical solutions:
[0023] A screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion, including:
[0024] An acquisition module, which is configured to acquire multi-channel data information of the personnel to be processed;
[0025] A screening module, which is configured to extract persons who have interrupted payment from persons to be processed based on the acquired multi-channel data information;
[0026] An analysis module is configured to analyze the data information of the proposed personnel who have interrupted payment to obtain a potential expansion database; analyze the obtained potential expansion database to obtain the uninsured data that can be expanded;
[0027] An extraction module is configured to extract the address information fields, address segmentation and coding classification of the obtained expandable uninsured data to obtain standard address information of expandable uninsured persons;
[0028] The positioning module is configured to compare and match the obtained standard address information with the medical insurance agency information in the local database, obtain the medical insurance agency to which the uninsured persons who can be expanded to cover the medical insurance, and complete the positioning of the insured persons who can be expanded to cover the medical insurance.
[0029] According to some embodiments, a third solution of the present invention provides a computer-readable storage medium, which adopts the following technical solution:
[0030] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps in the method for screening and locating persons participating in insurance coverage expansion based on multi-source data fusion as described in the first solution of the present invention.
[0031] According to some embodiments, a fourth solution of the present invention provides an electronic device, which adopts the following technical solution:
[0032] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the method for screening and locating people participating in insurance coverage expansion based on multi-source data fusion as described in the first scheme of the present invention.
[0033] According to some embodiments, a fifth solution of the present invention provides a computer program product, which adopts the following technical solution:
[0034] A computer program product includes software code, wherein the program in the software code executes the steps in the method for screening and locating people for insurance coverage expansion based on multi-source data fusion as described in the first solution of the present invention.
[0035] Compared with the prior art, the present invention has the following beneficial effects:
[0036] The present invention completes the coverage screening of the entire population by collecting and integrating multi-channel data and fusing multi-source data, thereby ensuring the comprehensiveness of data screening and effectively reducing the problem of inaccurate personnel screening caused by incomplete information or large errors from a single data source.
[0037] The present invention constructs models and optimizes them. Each model analyzes the expansion potential of a person separately. After comprehensively summarizing each model, the expansion potential of the person can be comprehensively evaluated, and the result data has high availability. By screening from a single model to comprehensive analysis of multiple models, the problem of difficulty in locating the insured expansion personnel and low accuracy of the expansion data in the traditional model can be effectively solved.
[0038] Through multi-layer data verification technology, the present invention can effectively eliminate false and invalid data, further reduce the target data of insured persons to be expanded, ensure the authenticity and accuracy of the expanded data, and reduce the verification pressure on business personnel.
[0039] The present invention realizes structured processing of address information, unifies data standards, improves address data reliability, and achieves accurate positioning of people participating in the insurance expansion through address information parsing and matching technology.
[0040] The present invention solves the inefficient mode of offline allocation by traditional handling personnel through scalable data allocation for uninsured persons. Through the automatic matching calculation of the system, the operation of manual offline matching allocation based on address information by the handling personnel is improved, the accuracy of data allocation is improved, and the workload of the handling personnel is reduced. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] The drawings in the specification that constitute a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments of this embodiment and their descriptions are used to explain this embodiment and do not constitute improper limitations on this embodiment.
[0042] Figure 1 This is a flow chart of a method for screening and locating insured persons based on multi-source data fusion in Embodiment 1 of the present invention;
[0043] Figure 2 This is a structural block diagram of the system for screening and positioning insured persons based on multi-source data fusion in the second embodiment of the present invention. DETAILED DESCRIPTION
[0044] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0045] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.
[0046] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0047] In the present invention, terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are relational words determined only for the convenience of describing the structural relationships of the various parts or elements of the present invention, and do not specifically refer to any part or element in the present invention and should not be understood as limitations on the present invention.
[0048] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.
[0049] Embodiment 1
[0050] Embodiment 1 of the present invention introduces a method for screening and locating insurance coverage expansion personnel based on multi-source data fusion.
[0051] like Figure 1 A screening and positioning method for insurance coverage expansion personnel based on multi-source data fusion is shown, comprising the following steps:
[0052] Step S1: Collect and process personnel’s local medical insurance data, external medical insurance data, and data shared by other functional departments;
[0053] Step S2: extract the personnel who have stopped paying premiums by collecting data, analyze the data characteristics of the personnel who have stopped paying premiums, build a corresponding expansion personnel screening model according to the business category, regularly screen out potential expansion data and store it in the database, marking it as fixed data;
[0054] Step S3: Verify the saved fixed data in multiple layers, remove the non-expandable data in turn, and finally obtain the valid expandable uninsured data;
[0055] Step S4: extracting address information fields and address information segmentation and coding, marking each address element with a unified code, and obtaining standard address information;
[0056] Step S5: Compare and match the administrative address information with the medical insurance agency information maintained in the database, locate the medical insurance agency to which the uninsured persons can be expanded, and distribute the data.
[0057] As one or more implementation modes, the process of collecting data information in step S1 is as follows:
[0058] Step S1-1: Collect local medical insurance data information, that is, collect incremental data on a daily basis through database tables, including basic personnel information, personnel insurance information, personnel payment information, unit information, school information, primary medical insurance zoning information, and benefit enjoyment information; specifically:
[0059] Basic personnel information includes data items such as ID number, registered residence, and permanent residence;
[0060] The insurance information of personnel includes data items such as insurance identity, insurance status, insurance start and end dates, insurance types, and interrupted insurance months. The insurance identities include employee insurance, flexible employment insurance, resident insurance, and interrupted insurance;
[0061] Personnel payment information includes employee payment year and month, employee payment period, employee payment type, flexible employment payment year and month, flexible employment payment period, flexible employment type, number of interrupted payment months, resident payment year, resident payment status, resident interruption year, etc.
[0062] Unit information includes data items such as unit number, unit name, unit address, etc.
[0063] School information includes data items such as school number, school name, and school address;
[0064] Primary medical insurance division information includes data items such as street name, district and county, community name, street, contact address, and contact phone number;
[0065] The information on benefit enjoyment includes data items such as the category of the beneficiary and the mark of the non-local personnel.
[0066] Step S1-2: Collect external medical insurance data information, including insurance information of other cities in the province, payment information of other cities in the province, and insurance information outside the province; specifically:
[0067] The provincial insurance payment information collection method is database table mode, and the collection cycle supports daily collection;
[0068] The method for collecting insurance information outside the province is interface mode, and the collection cycle can support collection triggered by half-month, month, quarter, etc. or function;
[0069] The out-of-province insurance information applies the query interface provided by the National Medical Insurance Information Platform, which supports single queries. In order to improve the efficiency of data query, the background scheduled task polling call is used to realize the download of out-of-province insurance information, and the data is redundantly stored in the local database to facilitate data query and use. A data periodic synchronization task is established to achieve monthly synchronization of the full amount of business data.
[0070] Step S1-3: Collect other data information, that is, collect shared data from other functional departments, support monthly or annual collection through database tables or files, including public security data, civil affairs data, health and health commission data, human resources and social security data, big data bureau data, etc.; specifically:
[0071] Public security data includes information on registered residents, floating residents, and deceased residents. The registered residents’ information includes data items such as ID number and registered address, the floating residents’ information includes data items such as ID number and permanent address, and the deceased’s information includes data items such as ID number, name, and date of death.
[0072] Civil affairs data mainly includes information about deceased persons, including ID number, ID type, name, date of death, and data source;
[0073] Funeral data include document number, document type, name, date of death, cremation date, and cremation address;
[0074] The data from the National Health Commission include certificate number, certificate type, name, contact number, and permanent residence;
[0075] Human resources and social security data include certificate number, certificate type, name, insurance address, household registration address, permanent address, contact number, insurance identity, and insurance status;
[0076] The data of the Big Data Bureau include certificate number, certificate type, name, contact number, and contact address;
[0077] For the death information shared by other functional departments, there is a problem of diversified data formats. Each data source has large differences in field types and field meanings. First, we established a unified death collection table for the province based on various death information tables, unified coding rules and field types, clearly recorded data sources, and verified each other through the intersection of death information from different sources, calculated the credibility of the death information, and provided it for later screening.
[0078] As one or more implementation modes, the specific process of step S2 is:
[0079] Step S2-1: Data analysis: by analyzing the personnel who have stopped paying premiums in the collected data, the common characteristics of such personnel are identified, and the target data is analyzed according to the change rules of the business data of the personnel who have stopped paying premiums;
[0080] Step S2-2: Model construction, classify the target data according to the insured status, age, benefit enjoyment, payment period, insurance interruption, and payment interruption, etc., set the priority of data screening according to the weight ratio of various influencing factors, and establish a classified expansion personnel screening model based on the above priorities and the cascade relationship between the data;
[0081] Step S2-3: Model optimization: Based on the model constructed in step S2-2, apply potential expansion data, analyze the results generated by each model, combine actual business data, and re-score the indicator weights of the model based on the dimension of high or low impact on insurance participation to improve the accuracy of model data screening;
[0082] Step S2-4: Data screening: set up regular data screening tasks and data execution plans based on incremental data (local medical insurance data, external medical insurance data, and data shared by other functional departments). Through the expandable personnel screening model, regularly extract the corresponding business data and mark it as potential expansion data;
[0083] Step S2-5: Save the result, format the potential expanded data, integrate the synonymous different code data into a unified standard coding format, and save it in the database to form fixed data.
[0084] People who have interrupted payment of premiums are divided into three categories: residents, employees, and flexible employment personnel; specifically, the business data characteristics of residents' interrupted payment include: no use of medical insurance reimbursement in recent years, younger age, low payment continuity, more interruptions in payment, and deaths at an older age; the business data characteristics of employees' interrupted payment include: frequent changes in units, being at the retirement age and not meeting the retirement period, and no use of medical insurance reimbursement in recent years; the business data characteristics of flexible employment's interrupted payment include: no use of medical insurance reimbursement in recent years, frequent interruptions in payment, and a low payment period.
[0085] This embodiment marks the above interrupted payment business data as target data, including fields such as insured identity, insurance status, interrupted insurance status, benefit enjoyment status, payment status, interrupted payment status, age, permanent residence, household registration place, insurance place, date of death, etc.
[0086] This embodiment takes personnel identity analysis as an example, and divides the obtained evaluation results into three levels: high, medium, and low based on the evaluation of the influence of insurance participation. The obtained personnel identity-influence factor weight table is shown in Table 1.
[0087] Table 1 Personnel identity-influencing factor weight table
[0088] Influencing factors\personnel identity Resident interruption Worker interruption Interruption of flexible employment age middle Low Low Cumulative reimbursement high Low high Insurance outage middle high middle Non-payment high Low high Payment period Low middle high
[0089] This embodiment is based on the evaluation made based on the success rate of remobilizing participants to participate in the insurance this year, and the evaluation results are divided into three levels: high, medium and low. The obtained personnel identity-age distribution weight table, personnel identity-cumulative reimbursement times weight table, personnel identity-interruption times weight table, personnel identity-interruption times weight table and personnel identity-payment years weight table are shown in Table 2, Table 3, Table 4, Table 5 and Table 6 respectively.
[0090] Table 2 Personnel identity-age distribution weight table
[0091] Age distribution\personnel identity Resident interruption Worker interruption Interruption of flexible employment Age < 18 high none none 18<=Age<40 middle middle middle 40<=Age<55 middle Low middle 55<=Age high Low Low
[0092] Table 3 Personnel identity-cumulative reimbursement times weight table
[0093]
[0094]
[0095] Table 4 Personnel identity-number of insurance interruption weight table
[0096] Number of insurance interruptions\personnel identity Resident interruption Worker interruption Interruption of flexible employment Number of times <2 middle high middle 2<=Number of times<4 middle middle middle 4<=number of times middle Low middle
[0097] Table 5 Personnel identity-Number of non-payment weights
[0098] Number of missed payments\personnel identity Resident interruption Worker interruption Interruption of flexible employment Number of times <2 high middle high 2<=Number of times<4 middle middle middle 4<=number of times Low middle Low
[0099] Table 6 Personnel identity-payment years weight table
[0100] Payment period\personnel status Resident interruption Worker interruption Interruption of flexible employment Years < 1 middle Low Low 1 < number of years < 10 middle middle middle 10 < years middle high high
[0101] In this embodiment, the initial weight score of the residents who have not enjoyed benefits in the resident model is relatively high. By comparing the expansion potential result data and the expandable feedback data, it is found that there is a large correlation between age and benefit enjoyment. In the fixed age groups of 2 and 3, the impact of medical benefit enjoyment is relatively large, and in the fixed age groups of 1 and 4, the impact is significantly reduced compared with the former. Therefore, combined with the calculation results, the weight score is re-evaluated, and the obtained age distribution-cumulative reimbursement number weight table is shown in Table 7, and the new weight score is applied to the model.
[0102] Table 7 Age distribution-weight table of cumulative reimbursement times
[0103]
[0104]
[0105] This embodiment discovers the proportion of indicator weights through regular data comparison and analysis, dynamically adjusts indicator scores based on actual conditions, improves model accuracy, and ensures that the model screens out personnel with a high probability of being expandable.
[0106] The data execution scheme in this embodiment, according to the incremental collection frequency of once a day, first associates the relevant tables, obtains the personnel changes in the recent time period according to the update time and partition time, and extracts the latest full data of the personnel based on the above-mentioned range of changed personnel, including basic information, insurance information, external medical insurance information, information shared by other departments, etc. Based on the set model and rules, the model is screened for such personnel, and finally the potential expansion data is obtained.
[0107] Synonymous data with different codes in this embodiment refers to data that express the same semantics but have different code values or names; the synonymous data with different codes are located, and the relationship between local values and provincial values is established based on the data connection. Such relationship is solidified into the program to achieve standardized processing of business data, obtain standardized result data with unified coding for the entire province, and save it in the database.
[0108] As one or more implementation modes, the specific process of step S3 is as follows:
[0109] Step S3-1: Compare the fixed data with the valid death data, mark the dead persons therein, and remove them from the fixed data to obtain the first process data;
[0110] Step S3-2: Verify the local insurance information, use the first process data to compare with the insurance data in the local medical insurance system, mark the people who have been insured locally, and remove them from the first process data to obtain the second process data;
[0111] Specifically, for employed workers who are insured this month, only whether their current insurance status is normal is determined; for those who are not insured this month, it is necessary to first determine whether their current insurance status is normal, and then determine whether there has been any payment within six months. Those who meet the above conditions will be identified as local insured persons; for retired employees, it is first determined whether their current insurance status is normal, and then whether the retirement account benefit mark is effective. Those who meet the above conditions will be identified as local insured persons; for local flexible employment personnel, it is first determined whether their current insurance status is normal, and then determine whether there has been any payment within twelve months. Those who meet the above conditions will be identified as local insured persons; for local residents, it is determined whether there is any resident payment or exemption within the current pooling year. Those who meet the above conditions will be identified as local insured persons.
[0112] Step S3-3: verify the insurance information of other cities in the province, compare the second process data with the medical insurance data of other cities in the province, mark the people who have been insured in other cities in the province, and remove them from the second process data to obtain the third process data;
[0113] Specifically, for employees, the data is first compared with the insurance data of employees in the province to determine whether the employee's insurance status in the province is normal, and those who meet the above conditions are identified as insured persons within the province; for residents, the data is first compared with the resident payment data of the province to determine the residents' maximum payment year and month or maximum exemption year and month, and those who meet the above conditions are identified as insured persons within the province; those who have been insured in other cities and prefectures in the province are marked and eliminated from the second process data to obtain the third process data.
[0114] Step S3-4: Verify the insurance information outside the province, compare the third process data with the insurance information outside the province, mark the persons who have been normally insured outside the province, and remove them from the third process data to obtain expandable data on uninsured persons.
[0115] As one or more implementation modes, in step S4, the address information field is first extracted, and the address information field is extracted through address information aggregation, data cleaning and rule definition; rules are set to prioritize each type of address, and address information extraction is completed by combining regular expressions and string matching, and address information fields are obtained through data cleaning, including medical insurance participation address, household registration address, permanent address, student address, etc.
[0116] In this embodiment, address information is aggregated, and relevant address information is extracted based on the expandable uninsured population data and the shared data of other functional departments, and the above data is saved as a set of personnel address information; based on the expandable uninsured personnel, the household registration information, permanent residence information, insured school information, insured unit information, and insured street community information of the local medical insurance are extracted; based on the shared data of other functional departments, the household registration information and permanent residence information are extracted.
[0117] The data cleaning in this embodiment is to clean the address information of the personnel, and obtain valid address data by removing invalid, duplicate or incorrectly formatted address data; specifically, remove redundant fields. The address information contains unnecessary fields. By presetting a collection of redundant fields, the data is parsed and compared to preliminarily remove such data; remove duplicate data. There may be duplicate records in multiple address information. By comparing the city suffix, district and county suffix, town and street suffix, etc. in the key fields, the duplicate data is located, the compared data is parsed, and the parsed address is compared with other address information and address library information. Finally, the data with a high degree of consistency is retained to remove redundant duplicate data, thereby realizing data deduplication.
[0118] The definition rules in this embodiment, namely, by defining data accuracy rules and data priority rules, obtain the best and most accurate personnel contact address information, and solve the personnel location matching problem in the case of multiple addresses; specifically, based on the data specification, the data accuracy rules are defined, the standard data format is defined first, and the data accuracy is verified by regular expressions;
[0119] Define standard data formats, such as:
[0120] Address information format: city-county-town-village, county-town-village, etc.;
[0121] School address information format: city-county-school name;
[0122] The format of the insured address information is: insured city-insured district and county-insured unit, etc.;
[0123] In this embodiment, the data accuracy is verified by regular expressions. The group data in the address information is first captured, including but not limited to cities, counties, districts, schools, towns, villages, offices, neighborhood committees, etc., and then the arrangement and combination of the group data in the address information is judged to determine whether it meets the address combination requirements, and then the accuracy of the data is determined and scored.
[0124] An example of a regular expression formula in this embodiment:
[0125] regex1: (? <province>[^Province]+Province|.+Autonomous region|[^Autonomous region]+Autonomous region|[^Autonomous region]+Autonomous region? <city>[^Autonomous Prefecture]+Autonomous Prefecture|[^City]+City|[^Public Security Bureau]+Public Security Bureau|.*? Region|.*? Administrative unit|.+League|City district? <county>[^City Center District]+City Center District|[^County]+County|[^Branch Bureau]+Branch Bureau|[^District]+District|[^City]+City? <village>[^street]+street|[^town]+town|[^office]+office|[^township]+township|[^police station]+police station? <town>[^community]+community|[^new village]+new village|[^village]+village).
[0126] regex2: (? <city>[^Autonomous Prefecture]+Autonomous Prefecture|[^City]+City|[^Public Security Bureau]+Public Security Bureau|.*? Region|.*? Administrative unit|.+League|City district? <county>[^City Center District]+City Center District|[^County]+County|[^Branch Bureau]+Branch Bureau|[^District]+District|[^City]+City? <village>[^street]+street|[^town]+town|[^office]+office|[^township]+township|[^police station]+police station? <town>[^community]+community|[^new village]+new village|[^village]+village).
[0127] regex3: (? <province>[^Province]+Province|.+Autonomous region|[^Autonomous region]+Autonomous region|[^Autonomous region]+Autonomous region? <county>[^City Center District]+City Center District|[^County]+County|[^Branch Bureau]+Branch Bureau|[^District]+District|[^City]+City? <village>[^street]+street|[^town]+town|[^office]+office|[^township]+township|[^police station]+police station? <town>[^community]+community|[^new village]+new village|[^village]+village).
[0128] regex4: (? <province>[^Province]+Province|.+Autonomous region|[^Autonomous region]+Autonomous region|[^Autonomous region]+Autonomous region? <city>[^Autonomous Prefecture]+Autonomous Prefecture|[^City]+City|[^Public Security Bureau]+Public Security Bureau|.*? Region|.*? Administrative unit|.+League|City district? <county>[^City Center District]+City Center District|[^County]+County|[^Branch Bureau]+Branch Bureau|[^District]+District|[^City]+City? <village>[^street]+street|[^town]+town|[^office]+office|[^township]+township|[^police station]+police station).
[0129] This embodiment defines priority rules based on a set of personnel address information, and performs analysis from three dimensions: data address integrity, update time, and data source. Specifically, data integrity analysis is judged from five dimensions: city, county, town, village, and house number. A weight is assigned to each address type, and cascade weights are assigned between different addresses. The weight value of data integrity is derived based on the analysis results.
[0130] For example, the institution address score: city (2 points), county (3 points), town (4 points), village (3 points), house number (2 points), cascade weight city and county (2 points), county town (4 points), town village (3 points), village house number (2 points), etc. Based on the analysis results, the institution weight table and cascade relationship weight table shown in Table 8 and Table 9 are obtained respectively:
[0131] Table 8 Institutional Weight Table
[0132] level Weight city Low county middle town high village middle House Number Low
[0133] Table 9 Cascade relationship weight table
[0134]
[0135]
[0136] The update time analysis in this embodiment defines weights based on the order of storage or update, and generally follows the principle that the newer the data, the higher the weight. For example, the first storage (2 points), the second storage (3 points), the third storage (4 points), etc., and the storage order weights shown in Table 10 are obtained:
[0137] Table 10: Weight table of storage order
[0138] Order of storage Weight First entry Low Second storage middle The third storage high
[0139] The data source analysis in this embodiment defines weights based on the credibility and efficiency of the data. The data source weight table obtained in this embodiment is shown in Table 11. For example, the public security household registration, medical insurance participation information, and human resources and social security participation information have higher weights, the public security permanent residence information and medical insurance permanent residence information have average weights, and the public security floating population has a lower weight.
[0140] Table 11 Data source weight table
[0141] Data Source Weight Public security personnel high Medical insurance address high Human Resources and Social Security Insurance Address high Public Security Resident Address middle Medical insurance permanent address middle Mobile police officers Low
[0142] In this embodiment, the weight values of different data sources will change with the application of the data. Through the accuracy of personnel positioning and personnel feedback, combined with the address information used, a comprehensive evaluation is made to determine the new weight ratio, realize self-correction of the rules, define priorities that are consistent with business facts, and solve the problem of personnel positioning matching in the case of multiple addresses.
[0143] The address segmentation and coding classification in this embodiment uses the forward maximum matching method (MM) and the reverse maximum matching method (RMM) to perform address segmentation to obtain address information, perform matching calculations based on address priority and integrity, obtain geographic element information, and classify valid address data using unified standard coding.
[0144] In this embodiment, address segmentation uses forward maximum matching (MM) and reverse maximum matching (RMM) methods to achieve address segmentation. First, data with high priority and accuracy are segmented to obtain more accurate address information fields, and the address information is decomposed into specific administrative agency names to facilitate the unified coding of address data in the later stage.
[0145] The address segmentation method defines the maximum word length N as the basic parameter for segmentation. Use the forward maximum matching method (MM): scan the address information from left to right, and take a string of length N each time to match it with the entries in the dictionary. If the match is successful, the string is segmented as a word; if the match fails, the string length is reduced and the match continues until the match is successful or the string length is 1. Use the reverse maximum matching method (RMM): scan the address information from right to left, and also use the maximum word length N for matching. By comparing the segmentation results of MM and RMM, the optimal segmentation result is selected as the final address segmentation.
[0146] In this embodiment, matching calculation is performed to obtain corresponding word segmentation information in order of priority and accuracy in the rules, and the address segmentation is matched from superior to subordinate in terms of regional level. The corresponding address information is obtained by fuzzy matching with the primary medical insurance zoning information. If no match is found, it is matched with the extended information of the institution name, and the matching range of the subordinate address is constrained by the superior address. That is, when the provincial level is Shandong Province, the subordinate cities will only be matched with the 16 cities in Shandong Province, and will no longer be matched with the city information in other provinces, thereby ensuring the correctness of the cascading relationship of the address information.
[0147] After obtaining the geographic elements of the address information, determine whether the address can be traced down to the town, street or community. If the goal is not achieved, new word segmentation information needs to be obtained, and the above steps are repeated to finally obtain the geographic elements of the address information.
[0148] In this embodiment, a unified standard coding classification is used to formulate a unified administrative address coding standard based on geographical elements. The geographical element information is coded and converted, and each word or phrase is converted into a unique code to achieve standardization of address information. According to the type and attributes of the address information, the coded administrative address information is classified to facilitate subsequent matching calculations based on priority and completeness.
[0149] As one or more implementation methods, step S5 matches the administrative address information of the uninsured persons who can be expanded with the medical insurance institution information maintained in the database, and screens out the medical insurance institution information that matches the administrative address information record, that is, locates the medical insurance institution to which the uninsured persons who can be expanded belong; and distributes the information of the uninsured persons who can be expanded to the corresponding medical insurance institution.
[0150] This embodiment completes the coverage screening of the entire population by collecting and integrating multi-channel data and fusing multi-source data, thereby ensuring the comprehensiveness of data screening and effectively reducing the problem of inaccurate personnel screening caused by incomplete information or large errors in a single data source; by building models and optimizing, each model analyzes the expansion potential of a person separately, and after comprehensively summarizing each model, the expansion potential of the person can be comprehensively evaluated, and the result data has high availability; from single-model screening to multi-model comprehensive analysis, the problem of difficulty in locating the insured expansion personnel and low accuracy of the expansion data in the traditional model can be effectively solved; through multi-layer data verification technology, it can be effective Effectively eliminate false and invalid data, further reduce the target data of insured persons, ensure the authenticity and accuracy of the expansion data, and reduce the verification pressure on business personnel; through the analysis and matching technology of address information, realize the structured processing of address information, unify data standards, improve the reliability of address data, and realize the precise positioning of insured persons; through the data allocation of uninsured persons who can be expanded, solve the inefficient mode of offline allocation of traditional handling personnel, and through the automatic matching calculation of the system, improve the operation of manual offline matching and allocation based on address information by handling personnel, improve the accuracy of data allocation, and reduce the workload of handling personnel.
[0151] Embodiment 2
[0152] Embodiment 2 of the present invention introduces a screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion.
[0153] like Figure 2 The system shown is a screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion, including a data collection module 101, a model building module 102, a data verification module 103, a data analysis and matching module 104, and a data allocation module 105.
[0154] As one or more implementation modes, the data collection module 101 is used to collect and process local medical insurance data, external medical insurance data, and data shared by other functional departments of personnel; the data collection module 101 includes a local medical insurance data module 1011, an external medical insurance data module 1012, and a death data module 1013.
[0155] In the local medical insurance data collection module 1011, incremental data is collected on a daily basis through the database table method, including basic personnel information, personnel insurance information, personnel payment information, unit information, school information, primary medical insurance zoning information, and benefit enjoyment information.
[0156] In the external medical insurance data collection module 1012, external medical insurance data collection is implemented, including insurance information of other cities and prefectures in the province, payment information of other cities and prefectures in the province, and insurance information outside the province.
[0157] In other data collection module 1013, shared data collection of other functional departments is realized, and monthly or annual data collection is supported through importing and collecting data through database tables or files, including public security data, civil affairs data, health and family planning commission data, human resources and social security data, big data bureau data, etc.
[0158] As one or more implementation modes, the model building module 102 is used to analyze the data characteristics of personnel who have interrupted payment, build a corresponding expandable personnel screening model according to the business category, regularly screen out potential expansion data, store it in the database, and mark it as fixed data; the model building module 102 includes a data analysis module 1021, a model building module 1022, a model optimization module 1023, a data screening module 1024, and a result saving module 1025.
[0159] In the data analysis module 1021, by analyzing the personnel who have interrupted payment in the collected data, the common characteristics of such personnel are identified, and the target data is analyzed based on the change pattern of the business data of the personnel who have interrupted payment.
[0160] In the model building module 1022, the target data is classified according to the insured status, age, benefit enjoyment, payment period, insurance lapse, and payment cessation, etc. The priority of data screening is set according to the weight ratio of each influencing factor, and based on the above priorities and the cascade relationship between the data, a classified expansion personnel screening model is established.
[0161] In the model optimization module 1023, based on the model constructed in 1022, potential expansion data is applied to analyze the results generated by each model, and combined with actual business data, the indicator weights of the model are re-scored based on the dimension of high and low impact on insurance participation to improve the accuracy of model data screening.
[0162] In the data screening module 1024, regular data screening tasks and data execution plans are set up based on incremental data (local medical insurance data, external medical insurance data, and data shared by other functional departments). Through the expandable personnel screening model, corresponding business data is regularly extracted and marked as potential expansion data.
[0163] In the result saving module 1025, the potential expanded data is formatted, the synonymous different code data are integrated into a unified standard coding format, and saved in the database to form fixed data.
[0164] As one or more implementation methods, the data verification module 103 is used to perform multi-layer data verification on the stored fixed data, eliminate the non-expandable data in turn, and finally obtain the valid expandable uninsured data. The data verification module 103 includes a death information comparison module 1031, a local insurance information verification module 1032, an insurance information verification module 1033 for other cities in the province, and an insurance information verification module 1034 for other cities outside the province.
[0165] In the death information comparison module 1031, the dead persons are marked and removed from the fixed data to obtain the first process data.
[0166] In the local insurance information verification module 1032, the first process data is compared with the insurance data in the local medical insurance system, and the persons who have been insured locally are marked and removed from the first process data to obtain the second process data.
[0167] In the other cities and prefectures in the province insurance information verification module 1033, the second process data is compared with the medical insurance data of other cities and prefectures in the province, and the persons who have been insured in other cities and prefectures in the province are marked and removed from the second process data to obtain the third process data.
[0168] In the out-of-province insurance information verification module 1034, the third process data is compared with the out-of-province insurance information, and the persons who have been normally insured outside the province are marked and eliminated from the third process data to obtain expandable data on uninsured persons.
[0169] As one or more implementations, the data parsing and matching module 104 is used to extract various types of address information from the expandable uninsured data, perform word segmentation, matching calculation and coding classification processing on the address information, and mark each address element with a unified code to obtain standard address information. The data parsing and matching module 104 includes an address information field extraction module 1041 and an address information word segmentation and coding classification module 1042.
[0170] In the address information field extraction module 1041, the address information field is extracted by address information aggregation, data cleaning and rule definition. Specifically, the address information field extraction module 1041 includes an address information aggregation module 1041-1, a field cleaning module 1041-2, and a rule definition module 1041-3.
[0171] In the address information aggregation module 1041-1, relevant address information is extracted based on the expandable data of uninsured people and the shared data of other functional departments, and the above data is saved as a personnel address information set.
[0172] In the field cleaning module 1041 - 2 , the personnel address information is cleaned to obtain valid address data by removing invalid, duplicate or incorrectly formatted address data.
[0173] In the rule definition module 1041-3, by defining data accuracy rules and data priority rules, the optimal and most accurate personnel contact address information is obtained to solve the personnel location matching problem in the case of multiple addresses.
[0174] In the address information segmentation and coding classification module 1042, the address segmentation and coding classification obtains geographical elements through address segmentation and matching calculation, and performs unified standard coding classification on the valid address data. Specifically, the address information segmentation and coding classification module 1042 includes an address segmentation module 1042-1, a matching calculation module 1042-2, and a unified standard coding classification module 1042-3.
[0175] In the address segmentation module 1042-1, the forward maximum matching method (MM) and the reverse maximum matching method (RMM) are used to implement the address segmentation and obtain the address information field. The address information is decomposed into specific administrative agency names to facilitate the unified coding of address data in the later stage.
[0176] In the matching calculation module 1042-2, the corresponding word segmentation information is obtained in order of priority and accuracy in the rules, and the address word segmentation is matched according to the regional level to obtain the geographical elements of the address information.
[0177] In the coding classification module 1042-3, a unified administrative address coding standard is formulated based on geographical elements, and the geographical element information of the address is coded and converted, and each word or phrase is converted into a unique code to achieve standardization of address information. The coded address information is classified according to the type and attribute of the address information.
[0178] As one or more implementations, the data distribution module 105 is used to compare and match the administrative address information with the medical insurance institution information maintained in the database, locate the medical insurance institution to which the insured person belongs, and distribute the data. The data distribution module 105 includes a data matching module 1051 and a data distribution module 1052.
[0179] In the data matching module 1051, the administrative address information of the uninsured persons who can be expanded is matched with the medical insurance institution information maintained in the database, and the medical insurance institution information that matches the administrative address information is screened out, that is, the medical insurance institution to which the insured persons who can be expanded belong is located.
[0180] In the data distribution module 1052, the information of uninsured persons who can be expanded is distributed to the medical insurance institutions to which they belong.
[0181] The detailed steps are the same as the method for screening and locating insured persons based on multi-source data fusion provided in Example 1, and will not be repeated here.
[0182] Embodiment 3
[0183] Embodiment 3 of the present invention introduces a screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion.
[0184] A screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion, including:
[0185] An acquisition module, which is configured to acquire multi-channel data information of the personnel to be processed;
[0186] A screening module is configured to extract the persons who have suspended payment from the persons to be processed based on the multi-channel data information obtained, and the extracted persons who have suspended payment are the screened target persons for insurance coverage expansion;
[0187] An analysis module is configured to analyze the data information of the proposed personnel who have interrupted payment to obtain a potential expansion database; analyze the obtained potential expansion database to obtain the uninsured data that can be expanded;
[0188] An extraction module is configured to extract the address information fields, address segmentation and coding classification of the obtained expandable uninsured data to obtain standard address information of expandable uninsured persons;
[0189] The positioning module is configured to compare and match the obtained standard address information with the medical insurance agency information in the multi-channel data information, obtain the medical insurance agency to which the person who has interrupted payment belongs, and complete the positioning of the people who have participated in the expansion of insurance coverage.
[0190] The detailed steps are the same as the method for screening and locating insured persons based on multi-source data fusion provided in Example 1, and will not be repeated here.
[0191] Embodiment 4
[0192] Embodiment 4 of the present invention provides a computer-readable storage medium.
[0193] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps in the method for screening and locating persons participating in insurance coverage expansion based on multi-source data fusion as described in the first embodiment of the present invention.
[0194] The detailed steps are the same as the method for screening and locating insured persons based on multi-source data fusion provided in Example 1, and will not be repeated here.
[0195] Embodiment 5
[0196] A fifth embodiment of the present invention provides an electronic device.
[0197] An electronic device includes a memory, a processor, and a program stored in the memory and running on the processor. When the processor executes the program, it implements the steps in the method for screening and locating people participating in insurance expansion based on multi-source data fusion as described in Example 1 of the present invention.
[0198] The detailed steps are the same as the method for screening and locating insured persons based on multi-source data fusion provided in Example 1, and will not be repeated here.
[0199] Embodiment 6
[0200] Embodiment 6 of the present invention provides a computer program product.
[0201] A computer program product includes software code, wherein the program in the software code executes the steps in the method for screening and locating persons participating in insurance coverage expansion based on multi-source data fusion as described in Example 1 of the present invention.
[0202] The detailed steps are the same as the method for screening and locating insured persons based on multi-source data fusion provided in Example 1, and will not be repeated here.
[0203] The above description is only a preferred embodiment of the present embodiment and is not intended to limit the present embodiment. For those skilled in the art, the present embodiment may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment shall be included in the protection scope of the present embodiment.< / village> < / county> < / city> < / province> < / town> < / village> < / county> < / province> < / town> < / village> < / county> < / city> < / town> < / village> < / county> < / city> < / province>
Claims
1. A method for screening and locating insured persons based on multi-source data fusion, characterized in that: include: Obtain multi-channel data information of personnel to be processed; Based on the multi-channel data information obtained, the personnel who have stopped paying are extracted from the personnel to be processed; Analyze the data of the people who have stopped paying premiums and obtain a potential expansion database; Analyze the potential expansion database to obtain the uninsured data that can be expanded; Extract the address information fields, address segmentation and coding classification of the expandable uninsured data to obtain the standard address information of the persons who have stopped paying premiums; The obtained standard address information is compared and matched with the medical insurance agency information in the multi-channel data information to obtain the medical insurance agency to which the uninsured persons who can be expanded in coverage belong, thereby completing the positioning of the insured persons who can be expanded in coverage.
2. A method for screening and locating insured persons based on multi-source data fusion as claimed in claim 1, characterized in that: Based on the multi-channel data information obtained, the persons who have interrupted payment are extracted from the persons to be processed; based on the data characteristics of the extracted persons who have interrupted payment, target data including at least the insured status, age, benefit enjoyment, payment period, insurance interruption and payment interruption are obtained; the obtained target data are classified, and the multi-channel data information obtained is screened regularly according to the target data and its weight to obtain a potential expansion database.
3. A method for screening and locating insured persons based on multi-source data fusion as claimed in claim 1, characterized in that: The address information fields of the expandable uninsured data are extracted through address information aggregation, data cleaning and rule definition, and the extracted address information fields are subjected to address segmentation and matching calculation to obtain the geographical elements of the address information fields; the obtained geographical elements of the address information fields are coded and converted to obtain the standard address information of the expandable uninsured persons.
4. A method for screening and locating insured persons based on multi-source data fusion as claimed in claim 1, characterized in that: The obtained standard address information is matched with the medical insurance institution information maintained in the local database, and the medical insurance institution information that matches the standard address information is screened out, that is, the medical insurance institutions to which the uninsured persons who can be expanded belong are obtained, and the medical insurance institutions to which the uninsured persons who can be expanded belong are located, thereby realizing the positioning of the insured persons who can be expanded.
5. A method for screening and locating insured persons based on multi-source data fusion as claimed in claim 1, characterized in that: The obtained potential expansion database is subjected to multi-layer data verification, and the non-expandable data in the potential expansion database are eliminated in turn to obtain the expandable but uninsured data.
6. A method for screening and locating insured persons based on multi-source data fusion as claimed in claim 1, characterized in that: The multi-channel data information obtained includes at least local medical insurance data, external medical insurance data and shared data.
7. A screening and positioning system for insurance coverage expansion personnel based on multi-source data fusion, characterized in that: include: An acquisition module, which is configured to acquire multi-channel data information of the personnel to be processed; A screening module, which is configured to extract persons who have interrupted payment from persons to be processed based on the acquired multi-channel data information; An analysis module is configured to analyze the data information of the proposed personnel who have interrupted payment to obtain a potential expansion database; analyze the obtained potential expansion database to obtain the uninsured data that can be expanded; An extraction module is configured to extract the address information fields, address segmentation and coding classification of the obtained expandable uninsured data to obtain standard address information of expandable uninsured persons; The positioning module is configured to compare and match the obtained standard address information with the medical insurance agency information in the multi-channel data information, obtain the medical insurance agency to which the uninsured persons who can be expanded to cover belong, and complete the positioning of the insured persons who can be expanded to cover.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method for screening and locating persons participating in the insurance expansion based on multi-source data fusion as described in any one of claims 1 to 6 are implemented.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the method for screening and locating people participating in the insurance expansion based on multi-source data fusion as described in any one of claims 1-6 are implemented.
10. A computer program product comprising software code, characterized in that The program in the software code executes the steps of the method for screening and locating people participating in the insurance expansion based on multi-source data fusion as described in any one of claims 1-6.
Citation Information
Patent Citations
Satellite remote sensing technology-based planting insurance precise underwriting method
CN112734579A
Spatial reasoning address back calculation positioning method
CN115292437A
Geographic entity positioning analysis method and device and computer equipment
CN116431625A
Social security data processing method and system based on big data
CN117314163A
Method, device, and equipment for test data screening, and storage medium
WO2020119114A1
Cited By
Dynamic configuration algorithm-based participation mobilization method, medium and equipment
CN120258635A
Multi-source data verification access method and system for uninsured persons
CN120494886A
A multi-source data verification and reach method and system for uninsured personnel
CN120494886B
