External data testing method, device, equipment and medium

Abstract

The invention discloses an external data testing method, device and system and a medium, and the method comprises the steps: obtaining an external data source according with a preset matching rule, and integrating the external data source and an internal data field into a test sample; cutting the test sample into a first test sample and a second test sample; matching the external data source with the first test sample, and adding successfully matched data into an attention list; constructing a prediction model according to the second test sample, and obtaining an alternative field set in the second test sample through an evaluation index; and testing an external data source according to the attention list and the alternative field set to obtain a data test report. By effectively evaluating and associating the external data source, the accuracy, integrity, timeliness and other elements of the data can be comprehensively evaluated, so that the quality and efficiency of accessing the external data are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, which is applied in finance and other scenarios, and in particular to an external data testing method, device, equipment and storage medium. Background Art

[0002] In the financial sector, data empowerment of business has become a consensus in recent years. In addition to internal company data and available shared data, access to external data can also effectively empower business. However, the quality of data provided by third-party data companies varies, and it is necessary to evaluate whether it meets the company's security and compliance requirements. In addition, the cost-effectiveness of data procurement also needs to be carefully considered. The traditional external data access process does not use a general process to solve these problems, and lacks a comprehensive evaluation of temporary external data sources, which is inconsistent with the current environment of cost reduction and efficiency improvement. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide an external data testing method, device, equipment and storage medium to solve the problems existing in the traditional external data access process, namely, the lack of comprehensive evaluation of data and the failure to effectively control the quality of external data.

[0004] In order to solve the above technical problems, the embodiment of the present invention provides an external data testing method, which adopts the following technical solution:

[0005] An external data testing method comprises the following steps:

[0006] Acquire an external data source that meets the preset matching rules, and integrate the external data source with the internal data field into a test sample;

[0007] cutting the test sample into a first test sample and a second test sample;

[0008] Matching the external data source with the first test sample, and adding successfully matched data to a watch list;

[0009] Building a prediction model according to the second test sample, and obtaining a candidate field set in the second test sample by using an evaluation index;

[0010] The external data source is tested according to the watch list and the candidate field set to obtain a data test report.

[0011] In one embodiment, the step of obtaining an external data source that meets a preset matching rule and integrating the external data source with an internal data field into a test sample includes:

[0012] Obtain the external data source, and determine whether the external data source meets the preset matching rules;

[0013] When the external data source meets the preset matching rule, obtaining the external data field;

[0014] Reviewing the external data field and the internal data field to determine whether the external data field and the internal data field meet the evaluation criteria;

[0015] When the external data field and the internal data field meet the evaluation criteria, the external data field and the internal data field are integrated into a test sample.

[0016] In one embodiment, when the external data field and the internal data field meet the evaluation criteria, integrating the external data field and the internal data field into a test sample includes:

[0017] Performing data preprocessing on the external data field and the internal data field;

[0018] Determine the mapping relationship between the external data field and the internal data field according to the mapping rule;

[0019] Integrate the external data fields and the internal data fields according to the mapping relationship to obtain a data field set;

[0020] A sampling algorithm is used to extract test samples from the data field set.

[0021] In one embodiment, matching the external data source with the first test sample and adding successfully matched data to a watch list includes:

[0022] Analyzing the first test sample and obtaining data features of the first test sample;

[0023] matching the list in the external data source with the data features;

[0024] When the list in the external data source successfully matches the data feature, the successfully matched list is analyzed, and the successfully matched watch list is added to the watch list according to the analysis result.

[0025] In one embodiment, matching the external data source with the first test sample and adding the successfully matched data to a watch list further includes:

[0026] When the list in the external data source fails to match the data feature, obtaining a collision rate between the list in the external data source and the data feature;

[0027] The database collision rate is compared with a preset database collision threshold. When the database collision rate is lower than the preset database collision threshold, the list of external data sources is re-acquired and a database collision is performed.

[0028] In one embodiment, after building a prediction model according to the second test sample and obtaining a candidate field set according to the second test sample through an evaluation index, the method further includes:

[0029] Performing a single factor analysis on the external data field to obtain an evaluation result;

[0030] Calculating statistical indicators on the external data fields to obtain indicator data;

[0031] Filtering candidate fields in the written external data fields according to the evaluation results and indicator data;

[0032] The screened candidate fields are added to the candidate field set.

[0033] In one embodiment, the testing of the external data source according to the attention list and the candidate field set and obtaining the data test report includes:

[0034] Get the total number of cases queried in the specified time in the history record system;

[0035] Obtaining old strategy hit cases and new strategy hit cases according to the data test report;

[0036] Calculate the total benefit value and the total loss value according to the total cases of the query, the old strategy hit cases and the new strategy hit cases;

[0037] Whether to receive an external data source is determined according to the total gain value and the total loss value.

[0038] In order to solve the above technical problems, the embodiment of the present invention also provides an external data testing device, which adopts the following technical solution:

[0039] An external data testing device, comprising:

[0040] A test sample module, used to obtain an external data source that meets the preset matching rules, and integrate the external data source with the internal data field into a test sample;

[0041] A sample segmentation module, used for segmenting the test sample into a first test sample and a second test sample;

[0042] A list matching module, used to match the external data source with the first test sample, and add the successfully matched data to a watch list;

[0043] A candidate field module, used for building a prediction model according to the second test sample, and obtaining a candidate field set in the second test sample through an evaluation index;

[0044] The data testing module is used to test the external data source according to the attention list and the candidate field set to obtain a data testing report.

[0045] In order to solve the above technical problems, an embodiment of the present invention further provides a computer device, which adopts the following technical solution:

[0046] A computer device comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the external data testing method described above when executing the computer-readable instructions.

[0047] In order to solve the above technical problems, an embodiment of the present invention further provides a computer-readable storage medium, which adopts the following technical solution:

[0048] A computer-readable storage medium stores computer-readable instructions, which implement the steps of the external data testing method described above when executed by a processor.

[0049] Compared with the prior art, the embodiments of the present invention mainly have the following beneficial effects:

[0050] The external data testing method described in the embodiment of the present invention obtains an external data source that meets the preset matching rules, integrates the external data source and the internal data field into a test sample; cuts the test sample into a first test sample and a second test sample; matches the external data source with the first test sample, and adds the successfully matched data to a watch list; builds a prediction model based on the second test sample, and obtains an alternative field set in the second test sample through evaluation indicators; tests the external data source according to the watch list and the alternative field set to obtain a data test report. The present invention can comprehensively evaluate the accuracy, completeness, timeliness and other factors of the data by effectively evaluating and associating the external data source, thereby improving the quality and efficiency of accessing external data. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the scheme in the present invention, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 is an exemplary system architecture diagram in which the present invention can be applied;

[0053] Figure 2 is a flow chart of an embodiment of an external data testing method according to the present invention;

[0054] Figure 3 is a structural schematic diagram of an embodiment of an external data testing device according to the present invention;

[0055] Figure 4 FIG. 1 is a schematic structural diagram of an embodiment of a computer device according to the present invention. DETAILED DESCRIPTION

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present invention; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention; the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present invention or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0057] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0058] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings.

[0059] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0060] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0061] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III, Moving Picture Experts Group Audio Layer 3), MP4 (Moving Picture Experts Group Audio Layer IV, Moving Picture Experts Group Audio Layer 4) players, laptop computers and desktop computers, etc.

[0062] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .

[0063] It should be noted that the external data testing method provided in the embodiment of the present invention is generally executed by a server / terminal device, and accordingly, the external data testing device is generally arranged in the server / terminal device.

[0064] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0065] Continue to refer Figure 2 , shows a flow chart of an embodiment of an external data testing method according to the present invention. The external data testing method comprises the following steps:

[0066] S1. Obtain an external data source that meets the preset matching rules, and integrate the external data source with the internal data field into a test sample;

[0067] S2, cutting the test sample into a first test sample and a second test sample;

[0068] S3, matching the external data source with the first test sample, and adding successfully matched data to a watch list;

[0069] S4, constructing a prediction model according to the second test sample, and obtaining a candidate field set in the second test sample through evaluation indicators;

[0070] S5. Test the external data source according to the attention list and the candidate field set to obtain a data test report.

[0071] In this example, we first obtain external data sources and determine whether these data sources meet the company's preset matching rules, and filter out qualified external data sources. The matching rules mainly evaluate the availability and empowerment potential of external data sources for the business. For example, when judging the operating risks of online ride-hailing vehicles, if the mileage data of online ride-hailing vehicles can be provided, the judgment effect of the model will be significantly improved.

[0072] Furthermore, the external data source is combined with the internal data field to form test samples, which are divided into two parts: the first test sample and the second test sample. The first test sample is used for blacklist analysis, while the second test sample is used for model building. The first test sample is used to match the external data source, and the successfully matched data, that is, the list data with corresponding items found in the external data source, will be added to the watch list.

[0073] Furthermore, the second test sample is used to build a prediction model, which will evaluate the predictive ability of the external data field on the business results. The performance of the model is measured by evaluation indicators (such as accuracy, KS statistic, AUC, etc.), and the fields that contribute the most to the model are identified to form a set of candidate fields.

[0074] Combined with the watchlist and candidate field sets, conduct a comprehensive test of the external data source, which may include the predictive ability of the model, the gain effect of the data, and the impact of data integration. Prepare a data test report based on the test results. The report will summarize the performance of the external data source, including its impact on business indicators, potential risks and benefits. The data test report provides the company with decision support on whether to adopt external data sources, helping to evaluate the long-term value and potential application scenarios of external data. Ensure the scientific evaluation and effective use of external data sources, improve the data-driven nature of business decisions, and optimize risk management strategies.

[0075] In one embodiment, the step of obtaining an external data source that meets a preset matching rule and integrating the external data source with an internal data field into a test sample includes:

[0076] Obtain the external data source, and determine whether the external data source meets the preset matching rules;

[0077] When the external data source meets the preset matching rule, obtaining the external data field;

[0078] Reviewing the external data field and the internal data field to determine whether the external data field and the internal data field meet the evaluation criteria;

[0079] When the external data field and the internal data field meet the evaluation criteria, the external data field and the internal data field are integrated into a test sample.

[0080] In this embodiment, data is obtained from external data providers or third-party sources. These external data sources may contain potential information that is valuable to the business. First, a preliminary evaluation is performed on the obtained external data sources to determine whether they are consistent with the company's business needs and goals. The evaluation is based on preset rules such as data relevance, availability, and compliance. When the external data source is considered to meet the compliance rules, we will further obtain the data fields therein.

[0081] Furthermore, external data fields and the company's existing internal data fields are reviewed to ensure that they meet the evaluation standards in terms of quality, format, completeness, etc. The evaluation standards may include key factors such as data accuracy, consistency, compliance, and security. Only when the fields of both data sources meet these standards are they suitable for integration for testing. Combine the external data fields that meet the evaluation standards with the internal data fields to create test samples. These samples will be used for subsequent data analysis, model building, and verification.

[0082] The integrated test samples are used to evaluate the impact of external data on business indicators, test the performance of models, and explore possible patterns and associations in the data. This ensures the effective use of external data sources and combines them with internal data to support business decisions. This helps improve the quality of data-driven decisions and optimize risk management and business processes.

[0083] In one embodiment, when the external data field and the internal data field meet the evaluation criteria, integrating the external data field and the internal data field into a test sample includes:

[0084] Performing data preprocessing on the external data field and the internal data field;

[0085] Determine the mapping relationship between the external data field and the internal data field according to the mapping rule;

[0086] Integrate the external data fields and the internal data fields according to the mapping relationship to obtain a data field set;

[0087] A sampling algorithm is used to extract test samples from the data field set.

[0088] In this embodiment, before integrating external and internal data fields, these data fields need to be preprocessed to ensure the quality of the data. The preprocessing steps may include cleaning the data (e.g., removing errors and duplicate records), standardizing the data format, processing missing values ​​and outliers, etc. These steps are crucial to ensure the accuracy of subsequent analysis.

[0089] Further, the correspondence between the external data fields and the internal data fields is determined according to the mapping rules. The mapping rules define how to correspond the external data fields to the internal data fields, which may be based on the name, content, purpose or other relevant attributes of the fields. Clear mapping rules help ensure the consistency and accuracy of data integration. According to these mapping relationships, the external data fields and the internal data fields are integrated to form a new data field set. Using these mapping relationships, the external data fields and the internal data fields are effectively integrated together to construct a data field set.

[0090] From the integrated data field set, test samples are extracted through sampling algorithms. Sampling algorithms can include simple random sampling, stratified sampling, systematic sampling, etc. The specific method selected should be determined based on the representativeness requirements of the sample and the characteristics of the data set. The extracted test samples will be used to build and verify the model, as well as to evaluate the impact of external data fields on business indicators. The quality and representativeness of the test samples have a direct impact on the performance and reliability of the model.

[0091] Specifically, determining the mapping relationship between external data fields and internal data fields first requires clarifying business needs and goals, understanding the types of data required and how they support business processes. Next, explore external data sources to understand their data structure, field types, data quality, and coverage.

[0092] Identify the fields in the external data and compare them with the fields in the internal data model. Based on the business logic and data characteristics, develop mapping rules, which may cover direct mapping (when the field names and meanings are the same), indirect mapping (when conversion or calculation is required), or composite mapping (when multiple external fields are mapped to one internal field). It is crucial to determine the key fields, such as user ID, product code, etc., which are usually used as the main basis for data association. Use data dictionaries and metadata to deeply understand the definition, format, and business meaning of the fields. Preprocess the external data fields, including format conversion, encoding unification, and missing value processing, to facilitate the smooth mapping process.

[0093] Use data mapping tools or software to assist in identifying and establishing mapping relationships between fields. Use data alignment techniques, such as fuzzy matching and regular expression matching, to solve the problem of inconsistency in field names. During the mapping process, continuously verify the accuracy of the mapping results to ensure data consistency and correctness. Establish a feedback mechanism to continuously adjust and optimize mapping rules based on mapping results and business feedback. Throughout the mapping process, follow data governance principles to ensure data quality and security. After the mapping relationship is determined, SQL, ETL tools, data integration platforms and other technical means can be used to achieve data integration to ensure that the mapping relationship is both accurate and meets business needs.

[0094] Selecting or generating test samples from the integrated data set usually involves the following sampling methods: Simple Random Sampling: The probability of each data point being selected is consistent. Sample selection is usually achieved through a random number generator. Stratified Sampling: The data set is divided into different subsets or "layers" that are defined based on a specific feature or variable. Samples are randomly selected from each layer to ensure that the proportion of samples in each layer matches the proportion in the population. Systematic Sampling: First, a sampling interval is set, and then data points are selected at fixed intervals after a randomly selected starting point. This method is suitable for situations where the data population is large and orderly. In practical applications, the selection of sampling methods depends on the characteristics of the data set, research objectives, resources, and time constraints. Appropriate sampling methods can enhance the reliability and validity of analytical results. When selecting samples, the size of the sample should also be considered to ensure that the sample size is representative of the entire data set and avoid waste of resources due to excessively large samples.

[0095] In one embodiment, matching the external data source with the first test sample and adding successfully matched data to a watch list includes:

[0096] Analyzing the first test sample and obtaining data features of the first test sample;

[0097] matching the list in the external data source with the data features;

[0098] When the list in the external data source successfully matches the data feature, the successfully matched list is analyzed, and the successfully matched watch list is added to the watch list according to the analysis result.

[0099] In this embodiment, the first test sample is analyzed to extract its data features, which include but are not limited to historical violations and fraud patterns, etc. By deeply analyzing the first test sample, key data features that are helpful for identifying and assessing risks can be extracted.

[0100] Furthermore, the list provided by the external data source is matched with the data features obtained from the first test sample to identify the list items that need special attention. When the external list matches the data features successfully, these successfully matched lists are then further analyzed to verify whether they really need to be added to the watch list. Based on the analysis results, if it is determined that the successfully matched lists belong to the high-risk or special attention categories, then these lists are added to the watch list for continuous monitoring. This process helps to improve the accuracy and efficiency of risk management.

[0101] Specifically, the decision-making process for determining whether to add a list from an external data source to a watch list includes determining the provider of the list, which can be the company itself (our side) or the external data source (the other party). For example, if the external data source (the other party) provides the list, they may provide 2,000 samples. The selection criteria are determined through negotiation to increase the collision rate, with the goal of matching at least 1,000 records with our data. In this step, it is necessary to consider the scope, volume, and specification differences of data between different companies, all of which may affect the final amount of matching data. The validity analysis of the successfully matched lists is conducted to evaluate the extent to which they complement the existing lists and to determine whether the accuracy of the lists is credible. If the list is provided by our side, we can provide 10,000 samples, and the external data source (the other party) will analyze whether these samples hit their database and conduct a list validity analysis. This process is designed to accurately identify and manage potential risks while ensuring the accuracy and validity of the watch list.

[0102] In one embodiment, matching the external data source with the first test sample and adding the successfully matched data to a watch list further includes:

[0103] When the list in the external data source fails to match the data feature, obtaining a collision rate between the list in the external data source and the data feature;

[0104] The database collision rate is compared with a preset database collision threshold. When the database collision rate is lower than the preset database collision threshold, the list of external data sources is re-acquired and a database collision is performed.

[0105] In this embodiment, when the list in the external data source fails to match the data features, this indicates that there is no need to add the list provided by the external data source to the watch list. Next, it is necessary to decide how the list is provided, that is, whether the company itself (our side) or the external data source (the other party) provides the list. Generally speaking, it is recommended that the company itself provide the list, so that a sufficient sample size can be retained to verify the effectiveness of the business strategy. However, if the list can only be provided by an external data source due to compliance requirements or other factors, the company needs to ensure that the collision rate is high enough to ensure that the data volume is statistically significant. The collision rate refers to the ratio of the amount of data remaining after the external list is matched with the company's internal data. If the collision rate is lower than the preset collision threshold, this indicates that the overlap between the external list and the company's existing data is not high, and the choice of the list may need to be reconsidered.

[0106] For example, if the company itself provides a list, it can prepare 1,000 reported cases as samples. Among these 1,000 cases, 100 may be referred for claim processing, while 15 may be rejected due to fraud, no responsibility, etc. When the external data source provides a list, it is necessary to provide 2,000 sample cases and all factors (fields) related to them. After matching with the company's internal data, it is assumed that 1,200 reported cases can be found, of which 110 are referred and 17 are rejected. Taking auto insurance claims as an example, referral refers to the claim processing after the report, including steps such as investigation and damage assessment; rejection refers to the situation where the claim is refused to pay based on reasons such as fraud, no responsibility, etc. The numbers provided here are all examples to show how to obtain and analyze external data by analyzing referrals and rejections.

[0107] Through this process, the company can evaluate the specific impact of factors (fields) provided by external data sources on claims cases and verify the effectiveness of business strategies. Even if the company decides not to add an external list to the watch list, it can verify the validity of external data and the applicability of business strategies by determining how the list is provided, analyzing the collision rate, and combining it with actual business cases.

[0108] In one embodiment, after building a prediction model according to the second test sample and obtaining a candidate field set according to the second test sample through an evaluation index, the method further includes:

[0109] Performing a single factor analysis on the external data field to obtain an evaluation result;

[0110] Calculating statistical indicators on the external data fields to obtain indicator data;

[0111] Filtering candidate fields in the written external data fields according to the evaluation results and indicator data;

[0112] The screened candidate fields are added to the candidate field set.

[0113] In this embodiment, the preliminary intention stage involves a comprehensive evaluation of external data sources to determine whether they are consistent with the company's business needs. Based on the results of the business fit assessment, external data fields that match the business goals and needs are screened out. Single factor analysis is performed on the screened external data fields to evaluate the independent impact of each field on business indicators (such as the accuracy of claim denial). Statistical methods such as saturation, IV (information value), and KS (Kolmogorov-Smirnov) statistics are used to measure the predictive ability of each field and its contribution to the model.

[0114] Furthermore, evaluate the gain of each variable to the model accuracy to determine whether there is a single variable that can significantly improve the predictive performance of the model. Based on the results of the univariate analysis, screen out the variables that have significant gains to the model, and these variables will be part of the candidate field set. If a single variable can significantly improve the model's claim rejection accuracy, such as an increase of more than 25%, the variable will be considered to have significant gains to the model and will be selected as a useful field. The screened variables will be used for subsequent model building and multivariate analysis to further optimize model performance. Univariate analysis is an iterative process that may need to be adjusted and optimized based on the results of model building. The results of the univariate analysis will support business decisions and help companies determine which external data fields are most valuable to the business.

[0115] Specifically, saturation is a statistic that measures the independent contribution of a variable in the prediction results. Generally, a saturation greater than 0.6 indicates that the variable has a high predictive ability. Information Value (IV) is another indicator to measure the predictive ability of a variable and is widely used in credit scoring models. Generally speaking, variables with IV greater than 0.02 are considered useful. The KS statistic (Kolmogorov-Smirnov Statistic) measures the difference between the cumulative distribution of good and bad samples and is an important indicator for evaluating the distinguishing ability of the model. A KS value greater than 0.05 usually indicates that the model has a certain degree of distinguishing ability.

[0116] In one embodiment, the testing of the external data source according to the attention list and the candidate field set and obtaining the data test report includes:

[0117] Get the total number of cases queried in the specified time in the history record system;

[0118] Obtaining old strategy hit cases and new strategy hit cases according to the data test report;

[0119] Calculate the total benefit value and the total loss value according to the total cases of the query, the old strategy hit cases and the new strategy hit cases;

[0120] Whether to receive an external data source is determined according to the total gain value and the total loss value.

[0121] In this embodiment, the history record system is used to obtain the total number of cases (T) queried within a specified time period, where T represents the total number of cases queried or processed within the specific time period. The total number of all processed cases within the same time period is obtained from the company's business record system. The company's reporting system is used to generate a report on the number of cases that meet specific query or review requirements. Relevant case data is extracted from the company's data warehouse and summarized to obtain the total number of cases. If the company adopts a case management system, statistical data of all pending or processed cases can be directly obtained from the system.

[0122] In the case of multiple data sources, information needs to be integrated to ensure the completeness and accuracy of the total number of cases. Forecasts are made by comparing historical data for the same period to estimate the total number of cases in the current time period. Based on the scope of the risk assessment, the types and number of cases to be evaluated are determined, which may include all claims cases, credit applications or other business cases. At the same time, based on the requirements of the regulator or compliance standards, the types and number of cases to be reviewed are determined. It is also necessary to define the scope of cases that require specific inquiries or analysis based on the specific needs of the business department.

[0123] In cases where it is not possible or practical to obtain data on all cases, statistical sampling methods may be used to estimate the total number of cases. Ensuring the accuracy and representativeness of the "total cases queried" data is critical to making effective cost-benefit analysis and production decisions.

[0124] According to the data testing report, obtain the hit cases of the old strategy (A) and the hit cases of the new strategy (B). Among them, A refers to the number of cases identified as high risk or requiring special treatment when using the original strategy or model. B refers to the number of cases identified as high risk or requiring special treatment after using the new strategy or model integrated with external data. Define the differences between the old and new strategies, including their respective rules, models, or standards. Prepare a test data set, which may cover historical case data or a dedicated test sample. Apply the old and new strategies to the same test data set, and record the number of cases hit by each strategy. The so-called hit case refers to the case that is identified as high risk or requires further review according to the strategy. Use the old strategy to screen the test data set, and the identified hit cases are recorded as the number of cases A. Use the new strategy to screen the same test data set, and the identified hit cases are recorded as the number of cases B.

[0125] Calculate the difference in claims costs between the new strategy and the old strategy, that is, the claims costs of the cases hit by the old strategy minus the claims costs of the cases hit by the new strategy (AB). Use the total number of cases T to calculate the query cost (Loss1), where Loss1 represents the cost of querying or reviewing all query cases. Get the average cost of handling each case and multiply this average cost by the total number of cases to get the total query cost.

[0126] Next, the cost of the promotion (Loss2) is calculated based on the difference in the number of cases hit by the new strategy and the old strategy (BA). Loss2 represents the additional promotion cost caused by the implementation of the new strategy. The average cost of promoting each additional case is determined, and this average cost is multiplied by the number of additional cases to obtain the additional promotion cost.

[0127] Further calculate the relationship between total gain (Gain) and total loss (Loss1+Loss2). Total gain refers to the cost saved by reducing claims, while total loss refers to the additional cost incurred by the implementation of the new strategy. If the implementation of the new strategy does not bring new gains, that is, the gain (Gain) is less than or equal to the total loss (Loss1+Loss2), the new strategy may not be adopted. Evaluate the long-term and short-term cost-effectiveness of the new strategy to ensure that the new strategy not only improves the accuracy of risk management but is also economically feasible. Consider the possible risk reduction and profit increase brought by the new strategy, and weigh them against the cost required to implement the new strategy. Based on the above analysis, the company will decide whether to adopt the new strategy. If the new strategy can bring net economic benefits and the risks are controllable, the new strategy may be adopted.

[0128] Further references Figure 3 , as a response to the above Figure 3 The present invention provides an embodiment of an external data test device, which is similar to Figure 3 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0129] like Figure 3 As shown, the external data testing device described in this embodiment includes: a test sample module 11, a sample segmentation module 12, a list matching module 13, a candidate field module 14 and a data testing module 15. Among them:

[0130] The test sample module 11 is used to obtain an external data source that meets the preset matching rules and integrate the external data source with the internal data field into a test sample;

[0131] A sample segmentation module 12, used for segmenting the test sample into a first test sample and a second test sample;

[0132] A list matching module 13, used to match the external data source with the first test sample, and add the successfully matched data to a watch list;

[0133] A candidate field module 14, configured to construct a prediction model according to the second test sample, and obtain a candidate field set in the second test sample by using an evaluation index;

[0134] The data testing module 15 is used to test the external data source according to the attention list and the candidate field set to obtain a data testing report.

[0135] like Figure 3 As shown, the test sample module 11, the sample segmentation module 12, the list matching module 13, the alternative field module 14 and the data test module 15 are connected in sequence. The module referred to in the present invention refers to a series of computer program instruction segments that can complete specific functions, which is more suitable for describing the external data test execution process than the program. For the specific implementation method of each module, please refer to the corresponding method embodiment above, which will not be repeated here.

[0136] In one embodiment, the test sample module 11 includes:

[0137] A matching judgment unit, used to obtain the external data source and judge whether the external data source meets the preset matching rules;

[0138] An external data field unit, used for obtaining an external data field when the external data source meets a preset matching rule;

[0139] An evaluation unit, used to review the external data field and the internal data field to determine whether the external data field and the internal data field meet the evaluation criteria;

[0140] The test sample unit is used to integrate the external data field and the internal data field into a test sample when the external data field and the internal data field meet the evaluation criteria.

[0141] In one embodiment, the test sample unit comprises:

[0142] A preprocessing subunit, used for performing data preprocessing on the external data field and the internal data field;

[0143] A mapping subunit, used to determine a mapping relationship between an external data field and an internal data field according to a mapping rule;

[0144] A data field set subunit, used to integrate the external data field and the internal data field according to the mapping relationship to obtain a data field set;

[0145] The sample sampling subunit is used to extract test samples from the data field set using a sampling algorithm.

[0146] In one embodiment, the list matching module 13 includes:

[0147] A data feature unit, configured to analyze the first test sample and obtain data features of the first test sample;

[0148] A list matching unit, used to match the list in the external data source with the data feature;

[0149] The successful matching unit is used to analyze the successfully matched list when the list in the external data source successfully matches the data feature, and add the successfully matched watch list to the watch list according to the analysis result.

[0150] In one embodiment, the list matching module 13 further includes:

[0151] A matching failure unit, used for obtaining a collision rate between the list in the external data source and the data feature when the list in the external data source fails to match the data feature;

[0152] The database collision unit is used to compare the database collision rate with a preset database collision threshold, and when the database collision rate is lower than the preset database collision threshold, re-acquire the list of external data sources and perform a database collision.

[0153] In one embodiment, after building a prediction model according to the second test sample and obtaining a candidate field set according to the second test sample through an evaluation index, the method further includes:

[0154] Performing a single factor analysis on the external data field to obtain an evaluation result;

[0155] Calculating statistical indicators on the external data fields to obtain indicator data;

[0156] Filtering candidate fields in the written external data fields according to the evaluation results and indicator data;

[0157] The screened candidate fields are added to the candidate field set.

[0158] In one embodiment, the external data testing device further includes:

[0159] Total cases unit, used to obtain the total cases queried in a specified time in the history record system;

[0160] A hit case unit, used for obtaining old strategy hit cases and new strategy hit cases according to the data test report;

[0161] A calculation unit, used for calculating a total benefit value and a total loss value according to the total cases of the query, the cases hitting the old strategy and the cases hitting the new strategy;

[0162] The data receiving unit is used to determine whether to receive an external data source according to the total gain value and the total loss value.

[0163] In the above embodiment, the present invention discloses an external data testing device, which belongs to the field of data processing technology. The present invention obtains an external data source that meets the preset matching rules, integrates the external data source and the internal data field into a test sample; cuts the test sample into a first test sample and a second test sample; matches the external data source with the first test sample, and adds the successfully matched data to a watch list; builds a prediction model based on the second test sample, and obtains an alternative field set in the second test sample through evaluation indicators; tests the external data source according to the watch list and the alternative field set to obtain a data test report. The present invention can comprehensively evaluate the accuracy, completeness, timeliness and other factors of the data by effectively evaluating and associating the external data source, thereby improving the quality and efficiency of accessing external data.

[0164] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0165] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0166] To solve the above technical problems, the embodiment of the present invention also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0167] The computer device 8 includes a memory 8a, a processor 8b, and a network interface 8c that are interconnected through a system bus. It should be noted that the figure only shows a computer device 8 having components 8a-8c, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital processors (DSPs), embedded devices, etc.

[0168] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0169] The memory 8a includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 8a can be an internal storage unit of the computer device 8, such as a hard disk or memory of the computer device 8. In other embodiments, the memory 8a can also be an external storage device of the computer device 8, such as a plug-in hard disk equipped on the computer device 8, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Of course, the memory 8a can also include both the internal storage unit of the computer device 8 and its external storage device. In this embodiment, the memory 8a is usually used to store the operating system and various application software installed on the computer device 8, such as a computer-readable instruction of an external data testing method, etc. In addition, the memory 8a can also be used to temporarily store various types of data that have been output or are to be output.

[0170] The processor 8b may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other external data test chip in some embodiments. The processor 8b is generally used to control the overall operation of the computer device 8. In this embodiment, the processor 8b is used to run the computer-readable instructions or process data stored in the memory 8a, such as the computer-readable instructions for running the external data test method.

[0171] The network interface 8c may include a wireless network interface or a wired network interface. The network interface 8c is generally used to establish a communication connection between the computer device 8 and other electronic devices.

[0172] The computer device proposed in this embodiment belongs to the field of data processing technology and is used in scenarios such as financial therapy. The present invention obtains an external data source that meets the preset matching rules, integrates the external data source and the internal data field into a test sample; cuts the test sample into a first test sample and a second test sample; matches the external data source with the first test sample, and adds the successfully matched data to a watch list; builds a prediction model based on the second test sample, and obtains an alternative field set in the second test sample through evaluation indicators; tests the external data source according to the watch list and the alternative field set to obtain a data test report. The present invention can comprehensively evaluate the accuracy, completeness, timeliness and other factors of the data by effectively evaluating and associating the external data source, thereby improving the quality and efficiency of accessing external data.

[0173] The present invention also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by a processor to enable the processor to perform the steps of the external data testing method as described above.

[0174] The computer-readable storage medium proposed in this embodiment belongs to the field of data processing technology and is applied to scenarios such as financial therapy. The present invention obtains an external data source that meets the preset matching rules, integrates the external data source and the internal data field into a test sample; cuts the test sample into a first test sample and a second test sample; matches the external data source with the first test sample, and adds the successfully matched data to a watch list; builds a prediction model based on the second test sample, and obtains an alternative field set in the second test sample through evaluation indicators; tests the external data source according to the watch list and the alternative field set to obtain a data test report. The present invention can comprehensively evaluate the accuracy, completeness, timeliness and other factors of the data by effectively evaluating and associating the external data source, thereby improving the quality and efficiency of accessing external data.

[0175] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0176] It should be noted that if software tools or components other than those of the company appear in the embodiments of the present application, they are only used for example introduction and do not represent actual use. The embodiments described above are only some embodiments of the present invention, not all embodiments. The preferred embodiments of the present invention are given in the accompanying drawings, but they do not limit the scope of the invention. The present invention can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive. Although the present invention is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific embodiments, or to replace some of the technical features therein with equivalents. Any equivalent structure made using the contents of the specification and drawings of the present invention, directly or indirectly used in other related technical fields, is also within the scope of patent protection of the present invention.

Claims

1. An external data testing method, characterized in that: The steps include: Acquire an external data source that meets the preset matching rules, and integrate the external data source with the internal data field into a test sample; cutting the test sample into a first test sample and a second test sample; Matching the external data source with the first test sample, and adding successfully matched data to a watch list; Building a prediction model according to the second test sample, and obtaining a candidate field set in the second test sample by using an evaluation index; The external data source is tested according to the watch list and the candidate field set to obtain a data test report.

2. The external data testing method according to claim 1, characterized in that: The obtaining of an external data source that meets the preset matching rules and integrating the external data source with the internal data field into a test sample includes: Obtain the external data source, and determine whether the external data source meets the preset matching rules; When the external data source meets the preset matching rule, obtaining the external data field; Reviewing the external data field and the internal data field to determine whether the external data field and the internal data field meet the evaluation criteria; When the external data field and the internal data field meet the evaluation criteria, the external data field and the internal data field are integrated into a test sample.

3. The external data testing method according to claim 2, characterized in that: When the external data field and the internal data field meet the evaluation criteria, the external data field and the internal data field are integrated into a test sample, including: Performing data preprocessing on the external data field and the internal data field; Determine the mapping relationship between the external data field and the internal data field according to the mapping rule; Integrate the external data fields and the internal data fields according to the mapping relationship to obtain a data field set; A sampling algorithm is used to extract test samples from the data field set.

4. The external data testing method according to claim 3, characterized in that: The step of matching the external data source with the first test sample and adding successfully matched data to a watch list includes: Analyzing the first test sample and obtaining data features of the first test sample; matching the list in the external data source with the data features; When the list in the external data source successfully matches the data feature, the successfully matched list is analyzed, and the successfully matched watch list is added to the watch list according to the analysis result.

5. The method for external data testing according to claim 4, characterized in that: The step of matching the external data source with the first test sample and adding successfully matched data to a watch list further includes: When the list in the external data source fails to match the data feature, obtaining a collision rate between the list in the external data source and the data feature; The database collision rate is compared with a preset database collision threshold. When the database collision rate is lower than the preset database collision threshold, the list of external data sources is re-acquired and a database collision is performed.

6. The external data testing method according to claim 2, characterized in that: After the prediction model is constructed according to the second test sample and a candidate field set is obtained through the evaluation index in the second test sample, the method further includes: Performing a single factor analysis on the external data field to obtain an evaluation result; Calculating statistical indicators on the external data fields to obtain indicator data; Filtering candidate fields in the written external data fields according to the evaluation results and indicator data; The screened candidate fields are added to the candidate field set.

7. The external data testing method according to claim 1, characterized in that: After testing the external data source according to the attention list and the candidate field set and obtaining the data test report, the method includes: Get the total number of cases queried in the specified time in the history record system; Obtaining old strategy hit cases and new strategy hit cases according to the data test report; Calculate the total benefit value and the total loss value according to the total cases of the query, the old strategy hit cases and the new strategy hit cases; Whether to receive an external data source is determined according to the total gain value and the total loss value.

8. An external data testing device, characterized in that: include: A test sample module, used to obtain an external data source that meets the preset matching rules, and integrate the external data source with the internal data field into a test sample; A sample segmentation module, used for segmenting the test sample into a first test sample and a second test sample; A list matching module, used to match the external data source with the first test sample, and add the successfully matched data to a watch list; A candidate field module, used for building a prediction model according to the second test sample, and obtaining a candidate field set in the second test sample through an evaluation index; The data testing module is used to test the external data source according to the attention list and the candidate field set to obtain a data testing report.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the external data testing method according to any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the external data testing method according to any one of claims 1 to 7 are implemented.

Images