Data evaluation method, system and device, storage medium and computer program product
Through automatic data extraction, encrypted transmission and visual interactive interface, the problem of high data security and technical thresholds for financial institutions in the evaluation of the value of three-party data products is solved, efficient and accurate data evaluation is achieved, labor costs are reduced and knowledge is supported. shared.
Patent Information
- Application Number
- CN202411006655.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-29
AI Technical Summary
When evaluating the value of three-party data products, financial institutions face problems such as inability to guarantee data security, high technical threshold, low evaluation efficiency, poor accuracy and inability to trace back. Especially traditional manual coding methods lead to sensitive data leakage, complex operations, and uneven code quality. All-in-one knowledge sharing is inefficient.
Through automatic data extraction, encrypted transmission and evaluation result generation, data extraction and report generation is used to utilize a visual interactive interface, providing automated data acquisition, processing and transmission processes, reducing technical thresholds, and through scalable data value evaluation algorithm templates Library and report templates implement consistency evaluation.
Effectively ensure the security and efficiency of data acquisition, processing and transmission, reduce labor costs, avoid sensitive data leakage, improve evaluation accuracy and efficiency, and support knowledge sharing and audit requirements.
Smart Images

Figure CN119048219A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer application technology, and more specifically, to a data evaluation method, system, device, storage medium and computer program product. Background Art
[0002] With the rapid development of financial services, financial institutions have an increasing demand for automated loan approval and risk prevention and control through big data technology. At present, there are many types of third-party data products on the market with varying quality. Financial institutions face huge challenges in evaluating the value and effectiveness of third-party data products from cooperative suppliers. Specifically, the entire process of traditional third-party data product value assessment is as follows: Figure 1 As shown in the figure, the traditional tripartite data value assessment method relies on manual coding assessment, which has the following problems:
[0003] 1. Data security cannot be guaranteed
[0004] The value assessment of data products involves the processing of a large amount of sensitive data information of sample customers. The traditional method of manual data extraction and email file interaction is very likely to cause sensitive data leakage.
[0005] 2. High technical threshold
[0006] The traditional manual coding method of third-party data value assessment requires data product value assessors to be proficient in Hadoop data processing, model development languages such as Python / Sas / R, value assessment indicators such as KS / Lift / IV, model operation, model report writing and other skills in order to be competent for data product value assessment work.
[0007] 3. Low evaluation efficiency
[0008] For each data product value assessment task, it is necessary to span multiple operating system platforms, apply multiple software tools, develop data extraction, debug model evaluation code, and write model test reports. The operation process is complicated and the assessment efficiency is extremely low.
[0009] 4. Poor evaluation accuracy
[0010] Since the quality of codes developed by different people varies, low-quality codes may lead to deviations in the value assessment of data products.
[0011] 5. The evaluation process cannot be reversed
[0012] The manual coding evaluation method leaves no trace of the entire process, and the data product evaluation process cannot meet internal and external audit requirements.
[0013] 6. Traditional evaluation methods are inefficient
[0014] The traditional method of each person manually conducting data product evaluation is not conducive to the accumulation and sharing of internal knowledge, and once personnel changes occur, knowledge gaps are very likely to occur. Summary of the invention
[0015] In order to solve the problems existing in the above-mentioned prior art or at least part of the problems, the embodiments of the present invention provide a data evaluation method, system, device, storage medium and computer program product, which can effectively ensure the security and efficiency of data collection, processing, transmission and storage through automatic data extraction, encrypted transmission, evaluation result generation and evaluation report generation.
[0016] According to the first aspect of the present invention, an embodiment of the present invention provides a data evaluation method, which includes: automatically calculating sample customer information in a business database according to a preset collection frequency, generating a sample customer information wide table according to the sample customer information, and automatically pushing the sample customer information wide table to a customer information library of a third-party data management platform for storage; obtaining data extraction information through a visual interactive interface, automatically extracting sample data files from the customer information library according to the data extraction information, and storing them in a test sample sampling library; automatically compressing and encrypting the sample data files and pushing them to a data supplier operation end, and synchronously storing them in the test sample sampling library; the data supplier operation end decrypts the compressed and encrypted sample data files, generates a sample feedback data file according to the decrypted sample data files, and stores the sample feedback data file in a test sample feedback library; generating evaluation result data according to a data value evaluation algorithm template and the sample feedback data file; generating a test evaluation report according to a data value evaluation report template and the evaluation result data.
[0017] According to the above-mentioned embodiment of the present invention, the security and efficiency of data collection, processing, transmission and storage can be effectively guaranteed through automatic data extraction, encrypted transmission, evaluation result generation and evaluation report generation; data extraction information is obtained through a visual interactive interface and sample data files are automatically extracted, so that evaluators do not need to master advanced model development and algorithm skills, thereby saving labor costs.
[0018] In some implementations of the present invention, automatically extracting the sample data file from the customer information database according to the data extraction information includes: encrypting the customer sensitive data in the sample data file and then extracting it.
[0019] According to the above-mentioned implementation mode of the present invention, by automatically encrypting and then extracting the customer's sensitive data, the problem of sensitive data leakage caused by manual data extraction and email interaction can be avoided.
[0020] In some embodiments of the present invention, storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data file with the sample data file in the test sample sampling library based on the customer ID number identification and / or mobile phone identification of the sample customer.
[0021] In some embodiments of the present invention, the data value assessment algorithm template is obtained through an extensible data value assessment algorithm template library, and the extensible data value assessment algorithm template library includes: a value assessment algorithm template for scoring data, a value assessment algorithm template for variable data, and a value assessment algorithm template for abnormal list data; the assessment result data includes: a sample data overview, coverage indicators, stability indicators, effectiveness indicators, correlation indicators, hit rate, accuracy rate, and recall rate indicators; generating a test evaluation report according to the data value assessment report template and the assessment result data includes: automatically filling the assessment result data into the data value assessment report template to generate a test evaluation report.
[0022] According to the above-mentioned implementation mode of the present invention, data evaluation is performed according to the value evaluation algorithm template of scoring data, variable data, and abnormal list data, so as to achieve consistent data evaluation and avoid the problem that the traditional evaluation method affects the accuracy of data product evaluation due to differences in code writing.
[0023] According to the second aspect of the present invention, an embodiment of the present invention provides a data evaluation system, which includes: a sample customer information calculation module, which is used to automatically calculate the sample customer information in the business database according to a preset collection frequency, generate a sample customer information wide table according to the sample customer information, and automatically push the sample customer information wide table to the customer information library of the third-party data management platform for storage; a sample data extraction module, which is used to obtain data extraction information through a visual interactive interface, automatically extract sample data files from the customer information library according to the data extraction information, and store them in a test sample sampling library; the sample data extraction module is also used to automatically compress and encrypt the sample data files and push them to the data supplier operation end, and synchronously store them in the test sample sampling library, the data supplier operation end decrypts the compressed and encrypted sample data files, generates a sample feedback data file according to the decrypted sample data files, and stores the sample feedback data files in the test sample feedback library; an evaluation result generation module, which is used to generate evaluation result data according to a data value evaluation algorithm template and the sample feedback data file; an evaluation report generation module, which is used to generate a test evaluation report according to a data value evaluation report template and the evaluation result data.
[0024] According to the above-mentioned embodiment of the present invention, the security and efficiency of data collection, processing, transmission and storage can be effectively guaranteed through automatic data extraction, encrypted transmission, evaluation result generation and evaluation report generation; data extraction information is obtained through a visual interactive interface and sample data files are automatically extracted, so that evaluators do not need to master advanced model development and algorithm skills, thereby saving labor costs.
[0025] In some implementations of the present invention, automatically extracting the sample data file from the customer information database according to the data extraction information includes: encrypting the customer sensitive data in the sample data file and then extracting it.
[0026] According to the above-mentioned implementation mode of the present invention, by automatically encrypting and then extracting the customer's sensitive data, the problem of sensitive data leakage caused by manual data extraction and email interaction can be avoided.
[0027] In some embodiments of the present invention, storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data file with the sample data file in the test sample sampling library based on the customer ID number identification and / or mobile phone identification of the sample customer.
[0028] In some embodiments of the present invention, the data value assessment algorithm template is obtained through an extensible data value assessment algorithm template library, and the extensible data value assessment algorithm template library includes: a value assessment algorithm template for scoring data, a value assessment algorithm template for variable data, and a value assessment algorithm template for abnormal list data; the assessment result data includes: a sample data overview, coverage indicators, stability indicators, effectiveness indicators, correlation indicators, hit rate, accuracy rate, and recall rate indicators; generating a test evaluation report according to the data value assessment report template and the assessment result data includes: automatically filling the assessment result data into the data value assessment report template to generate a test evaluation report.
[0029] According to the above-mentioned implementation mode of the present invention, data evaluation is performed according to the value evaluation algorithm template of scoring data, variable data, and abnormal list data, so as to achieve consistent data evaluation and avoid the problem that the traditional evaluation method affects the accuracy of data product evaluation due to differences in code writing.
[0030] According to the third aspect of the present invention, an embodiment of the present invention provides a computer-readable storage medium having computer-readable instructions stored thereon. When the computer-readable instructions are executed by a processor, the computer performs the following operations: the operations include the steps included in the data evaluation method described in any of the above embodiments.
[0031] According to the fourth aspect of the present invention, an embodiment of the present invention provides a computer device including a memory and a processor, wherein the memory is used to store one or more computer-readable instructions, wherein the one or more computer-readable instructions, when executed by the processor, can implement the data evaluation method described in any one of the above embodiments.
[0032] According to a fifth aspect of the present invention, an embodiment of the present invention provides a computer program product comprising a computer program, and when the computer program is executed by a processor, the data evaluation method as described in any one of the above embodiments is implemented.
[0033] From the above, it can be seen that the data evaluation method, system, device, storage medium and computer program product provided in the embodiments of the present invention can effectively ensure the security and efficiency of data collection, processing, transmission and storage through automatic data extraction, encrypted transmission, evaluation result generation and evaluation report generation; obtain data extraction information and automatically extract sample data files through a visual interactive interface, so that evaluators do not need to master advanced model development and algorithm skills, thereby reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 It is a flowchart of the traditional third-party data product value assessment method;
[0035] Figure 2 is a schematic flow chart of a data evaluation method according to Embodiment 1 of the present invention;
[0036] Figure 3 is a schematic diagram of a sample customer information wide table according to an embodiment of the present invention;
[0037] Figure 4 is a schematic diagram of a visual interactive interface according to an embodiment of the present invention;
[0038] Figure 5 is a schematic diagram of a data value assessment report template according to an embodiment of the present invention;
[0039] Figure 6 is a schematic diagram of a test evaluation report according to an embodiment of the present invention;
[0040] Figure 7 is a schematic diagram of the architecture of an Internet distributed cluster technology for implementing the data evaluation method of Example 1 according to an embodiment of the present invention;
[0041] Figure 8 2 is a schematic diagram of the architecture of a data evaluation system according to Embodiment 2 of the present invention. DETAILED DESCRIPTION
[0042] The various aspects of the present invention are described in detail below in conjunction with the accompanying drawings and specific embodiments. Among them, well-known modules, units and their connections, links, communications or operations are not shown or described in detail. In addition, the described features, architectures or functions can be combined in any way in one or more embodiments. It should be understood by those skilled in the art that the various embodiments described below are only used for illustration and are not intended to limit the scope of protection of the present invention. It can also be easily understood that the modules or units or processing methods in each embodiment described herein and shown in the accompanying drawings can be combined and designed in various different configurations.
[0043] The following is a brief description of the terms used in this article.
[0044] Airflow: Apache Airflow, an open source workflow automation and scheduling platform, is designed to define, schedule, and monitor complex data pipelines, workflows, and automated tasks through a directed acyclic graph (DAG).
[0045] Kettle: KDE Extraction, Transportation, Transformation and Loading Environment, an open source ETL (Extract-Transform-Load) tool.
[0046] Portal: An application framework that integrates various application systems, data resources and Internet resources into an information management platform and provides users with a unified user interface. A portal is the entrance for users to access the Internet. It is not only a website that integrates diverse content services, but is also called a network portal. It is also generally considered to be the starting point for Internet browsers.
[0047] CSV: (Comma Separated Values) file, a common plain text file format used to store tabular data.
[0048] Excel: A spreadsheet software developed by Microsoft and an important component of the Microsoft Office suite. It is mainly used for data organization, calculation, analysis and visualization. Users can create tables in Excel and enter various types of data in cells, such as text, numbers, dates, etc.
[0049] PSI: Sample stability index, an indicator used to measure the offset caused by sample changes, usually used to measure the stability of the sample.
[0050] KS: A discrimination index commonly used to evaluate the performance of a binary classification model. It measures the maximum distance between positive and negative samples.
[0051] Lift: It is the improvement index of the model, which is used to measure how much better a model is at predicting the "response" in the target than random selection. A lift greater than 1 means that the model captures more "responses" than random selection.
[0052] IV: refers to information value or information volume. It is an indicator used to measure the predictive power of independent variables on target variables.
[0053] Hadoop: Hadoop is an open source distributed computing platform. It mainly consists of two parts: HDFS (Hadoop Distributed File System) and MapReduce (Distributed Computing Framework). Hadoop can process large-scale data and has the characteristics of high reliability, high scalability and high efficiency. HDFS is used to store large-scale data and distribute the data on multiple nodes to achieve data redundancy and fault tolerance. MapReduce is used to process and calculate large-scale data in parallel, decompose complex tasks into multiple small tasks, and execute them simultaneously on multiple nodes, thereby improving computing efficiency.
[0054] [Example 1]
[0055] Figure 2 4 is a flow chart of a data evaluation method according to Embodiment 1 of the present invention.
[0056] like Figure 2 As shown, in Embodiment 1 of the present invention, the data evaluation method may include: step S11, step S12, step S13, step S14, step S15 and step S16, and the above steps are described in detail below.
[0057] In step S11, the sample customer information in the business database is automatically calculated according to the preset collection frequency, a sample customer information wide table is generated according to the sample customer information, and the sample customer information wide table is automatically pushed to the customer information library of the third-party data management platform for storage. In this embodiment, according to the preset collection frequency, the database stored procedure SQL code is used to automatically extract and summarize data from the business database to obtain a sample customer information wide table, and the sample customer information wide table is automatically pushed to the customer information library of the third-party data management platform for storage. Among them, the customer information wide table includes three parts: the first part is the customer's four-element information, including name, ID card, mobile phone number, bank card number, and this element information is the key field for data product suppliers to match data; the second part is the customer's loan information, including but not limited to one or more of the following: business type, channel code, credit date, loan date, loan due date, loan status, settlement status, etc. The loan information is the basic field used by financial institutions to screen test samples; the third part is the customer's overdue performance information, including the customer's maximum historical highest overdue days, the customer's maximum point-in-time overdue days, etc., which is the key field used by financial institutions to identify good and bad customers.
[0058] In an exemplary embodiment, the big data platform uses Airflow automatic task scheduling software to schedule and execute the sample customer information storage process daily to generate Figure 3 The sample customer information wide table is shown, and the wide table data is automatically pushed to the third-party data management platform through the Kettle tool.
[0059] In step S12, data extraction information is obtained through a visual interactive interface, and sample data files are automatically extracted from the customer information library according to the data extraction information, and stored in a test sample sampling library. In some embodiments, automatically extracting sample data files from the customer information library according to the data extraction information includes: encrypting the customer sensitive data in the sample data file and then extracting it. Exemplarily, encryption algorithms such as MD5 and SHA256 developed by built-in Java are selected for encryption of customer basic information (e.g., name, ID number, mobile phone number, bank card number, etc.), so that sensitive data encryption processing is automatically completed when sample data is extracted.
[0060] In an exemplary embodiment, data extraction information is determined by setting data conditions in a visual interactive interface, and sample data is automatically screened from the customer information library according to the set data conditions (data extraction information). Furthermore, a Y mark of good and bad customers is automatically generated in the sample data, and this indicator is used as the evaluation basis for the discrimination between good and bad samples of the data to be tested. Optionally, the Y mark of good and bad customers is defined according to the threshold of the specified overdue indicator, such as defining a bad customer as a customer with a historical maximum overdue number of days exceeding 60 days, and otherwise a good customer, wherein the third-party data management platform can automatically generate the Y mark of good and bad customers by setting the sample data volume and the ratio of good and bad customers (for example, the ratio of good and bad customer sample extraction is 9:1). At the same time, in order to ensure the security of customer data, it supports the use of encryption algorithms such as MD5 and SHA256 developed in Java for customer basic element information (name, ID number, mobile phone number, bank card number, etc.), and automatically completes sensitive data encryption processing when extracting sample data.
[0061] The present invention provides an example of setting data extraction conditions and performing data extraction through a visual interactive interface: Figure 4 As shown, the test data name, associated sample data, extraction matching fields (including customer name, mobile phone number, ID number, bank card number, credit date, etc.), encryption, encryption algorithm, field value range, good / bad customer fields and conditions, etc. are set through the visual interactive interface. Sample data is automatically screened from the customer information library according to the set data conditions (data extraction information). Further, the test is performed according to the associated sample data (test data) and the referenced test code template, and then the test report is generated according to the referenced test report template, so that the evaluator does not need to master advanced model development and algorithm skills, thereby saving labor costs.
[0062] In step S13, the sample data file is automatically compressed and encrypted and then pushed to the data supplier operation end, and is synchronously stored in the test sample sampling library.
[0063] In this embodiment, the third-party data management platform has a built-in data supplier management portal, which supports the creation of supplier accounts for external data suppliers, thereby realizing a collaborative work platform between financial institutions and data suppliers; the third-party data management platform also provides a data push function. After selecting the supplier of sample data to be received and the field to be pushed, the sample data generated in step S12 can be automatically compressed and encrypted by the Zip algorithm and pushed to the corresponding supplier of sample data to be received, realizing second-level transparent transmission, and leaving traces through transmission records. The trace data includes transmission time, data sender, data file name, and data recipient, so as to ensure data transmission security and efficiency, and meet internal and external audit requirements. Among them, the field to be pushed does not include the sensitive field of the good and bad customer identification Y variable. The good and bad customer identification Y variable is stored locally and is not displayed to the outside world. It is only used to verify the validity of data products within the financial institution. In the subsequent step S14, after the supplier returns the customer feedback data file, the platform can match it with the local original sample data according to the encrypted customer factor information, realizing the automatic integration of the good and bad customer identification Y variable and the feedback data, eliminating the tedious operation of manual offline matching.
[0064] In step S14, the data supplier operation terminal decrypts the compressed and encrypted sample data file, generates a sample feedback data file according to the decrypted sample data file, and stores the sample feedback data file in a test sample feedback library.
[0065] In a further implementation, the sample feedback data file is transmitted back to the financial institution account through the data supplier operation terminal, and the data in the file will be synchronously stored in the test sample feedback library in the platform. The sample feedback data file is in CSV or Excel format. After the financial institution end of the third-party data management platform receives the sample feedback data file, it automatically matches the original sample data and stores it in the test sample feedback library, which is convenient for subsequent data processing and statistics, and at the same time achieves the purpose of leaving traces.
[0066] In some embodiments, after the data supplier completes receiving and decrypting the sample data, it uses MD5 or SHA256 to match the customer's basic factor information with local third-party data products to generate a sample feedback data file, which may include the customer's risk label, value label, customer score and other information.
[0067] In a further embodiment, storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data with the sample data in the test sample sampling library based on elements such as the sample customer's ID number identification and / or mobile phone identification.
[0068] In this embodiment, the above steps S11 to S14 are performed by a visual interactive sample data extraction engine.
[0069] In step S15, evaluation result data is generated according to the data value evaluation algorithm template and the sample feedback data file.
[0070] In some embodiments, the data value assessment algorithm template is obtained through an extensible data value assessment algorithm template library, and the extensible data value assessment algorithm template library includes: a value assessment algorithm template for scoring data, a value assessment algorithm template for variable data, and a value assessment algorithm template for abnormal list data. Among them, the extensible data value assessment algorithm template library is a Python algorithm template library, which is developed using the Python language. Through the three common code templates of the above-mentioned common data types, 95% of daily data testing needs can be met. Furthermore, the extensible data value assessment algorithm template library has a private code template, which can meet 5% of personalized data testing needs.
[0071] In a further embodiment, the evaluation result data includes: sample data overview, coverage index, stability index, effectiveness index, relevance index, hit rate, accuracy rate, and recall rate index. The specific meanings of the output indexes are as follows:
[0072] (1) Sample data overview: an overall overview of the data, including information such as the ratio of good and bad samples.
[0073] (2) Coverage index: the ratio between the number of non-missing samples in the data set and the total number of samples studied.
[0074] (3) Stability index: The stability of a feature determines whether it can be used effectively in the long term, so it is necessary to select variables with higher stability as much as possible. For example, the stability can be scored using indicators such as quantile distribution, distribution graph display, and PSI value.
[0075] (4) Validity index: The validity index is used to detect whether the feature can predict the Y variable well. Its evaluation indicators include IV, KS, and Lift evaluation indicators.
[0076] (5) Correlation index: If the correlation of data sources is too high, it means that the information brought is similar. Therefore, while ensuring effectiveness, try to keep the correlation between features as low as possible.
[0077] (6) Hit rate, accuracy, and recall rate indicators are often used for exception list and rule data.
[0078] ① Hit rate = number of hits in the abnormal list / total number of hits;
[0079] ② Accuracy = the number of bad hits in the anomaly list / the number of hits, that is, the proportion of truly bad data in the provided anomaly list data;
[0080] ③ Recall rate = the number of abnormal list hits and bad / the number of truly bad, that is, how much of the abnormal list data can be captured in the bad samples. This indicator is consistent with the calculation of the effective difference rate.
[0081] In step S16, a test evaluation report is generated according to the data value evaluation report template and the evaluation result data. In some embodiments, generating a test evaluation report according to the data value evaluation report template and the evaluation result data includes: automatically filling the evaluation result data into the data value evaluation report template to generate a test evaluation report.
[0082] The present invention provides an example of a data value assessment report template, such as Figure 5 As shown in the figure, the data value assessment report template uses the Word format to embed macro variables to automatically fill the assessment result data into the report. For example, {{product}} represents the product name of this test, {{name}} represents the name of the tester, and {{department}} represents the name of the tester's department...
[0083] By automatically filling the evaluation result data into the data value evaluation report template, on the one hand, the technical threshold is very low, and only basic Word skills are needed to design a new template. On the other hand, the test report generated by Word is convenient for secondary modification, printing and reporting, and knowledge sharing. At the same time, through the data value evaluation template library and the data value evaluation report library, evaluation knowledge sharing is realized, which helps to improve the continuity of the work of the three parties on data value and the inheritability of knowledge. In the case of personnel changes, data evaluation business knowledge can also be quickly mastered by reading historical data value evaluation reports.
[0084] The present invention also provides an example of a generated test evaluation report, such as Figure 6 As shown, the test evaluation report includes: test background description, automated test virtual data product evaluation data description, automated test virtual data product evaluation situation, and evaluation conclusion; wherein, the automated test virtual data product evaluation situation includes: feature coverage, overall distribution of features, feature stability index, feature effectiveness index, and feature sorting.
[0085] In some embodiments, the above steps S15 and S16 are performed by an automated value assessment engine, which automatically performs data value assessment by loading supplier feedback data files, an extensible data value assessment Python algorithm template library, and an extensible data value assessment report Word template library, and generates a three-party data value assessment report in minutes.
[0086] In this embodiment, based on Figure 7 The Internet distributed cluster technology shown in FIG. 1 implements the above steps S11 to S16, as shown in FIG. Figure 7 As shown, the Internet distributed cluster technology includes: microservices, message queues, memory databases, ES search engine databases, etc. Relying on the visual interactive sample data extraction engine and the automated three-party data value assessment engine, the customer information data processing, sample data extraction, sample data transmission, return sample data storage, data product value assessment, and data value assessment report generation tasks are efficiently connected in series, achieving minute-level sample data extraction and data assessment report generation, greatly improving the efficiency of data product assessment and reducing the cost of data value assessment.
[0087] The data evaluation method of Example 1 of the present invention can effectively ensure the security and efficiency of data collection, processing, transmission, and storage through automatic data extraction, encrypted transmission, evaluation result generation, and evaluation report generation; data extraction information is obtained through a visual interactive interface and sample customer data is automatically extracted, so that evaluators do not need to master advanced model development and algorithm skills, thereby saving labor costs.
[0088] [Example 2]
[0089] Figure 8 2 is a schematic diagram of the architecture of a data evaluation system according to Embodiment 2 of the present invention.
[0090] like Figure 8 As shown, the data evaluation system includes: a sample customer information calculation module 210 , a sample data extraction module 220 , an evaluation result generation module 230 , and an evaluation report generation module 240 .
[0091] Among them, the sample customer information calculation module 210 is used to automatically calculate the sample customer information in the business database according to the preset collection frequency, generate a sample customer information wide table based on the sample customer information, and automatically push the sample customer information wide table to the customer information library of the third-party data management platform for storage.
[0092] The sample data extraction module 220 is used to obtain data extraction information through a visual interactive interface, automatically extract sample data files from the customer information library according to the data extraction information, and store them in the test sample sampling library. Wherein, sample customer data is automatically extracted from the customer information library according to the data extraction information, and the sample customer data constitutes a sample data file.
[0093] In some implementations, automatically extracting the sample data file from the customer information repository according to the data extraction information includes: encrypting the customer sensitive data in the sample data file and then extracting it.
[0094] The sample data extraction module 220 is also used to automatically compress and encrypt the sample data file and push it to the data supplier operation end, and synchronously store it in the test sample sampling library. The data supplier operation end decrypts the compressed and encrypted sample data file, generates a sample feedback data file based on the decrypted sample data file, and stores the sample feedback data file in the test sample feedback library.
[0095] In some embodiments, storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data file with the sample data file in the test sample sampling library according to the customer ID number identification and / or mobile phone identification of the sample customer.
[0096] The evaluation result generation module 230 is used to generate evaluation result data according to the data value evaluation algorithm template and the sample feedback data file.
[0097] In some embodiments, the data value assessment algorithm template is obtained through an extensible data value assessment algorithm template library, and the extensible data value assessment algorithm template library includes: value assessment algorithm templates for scoring data, value assessment algorithm templates for variable data, and value assessment algorithm templates for exception list data; the assessment result data includes: sample data overview, coverage index, stability index, effectiveness index, relevance index, hit rate, accuracy, and recall rate index.
[0098] The evaluation report generation module 240 is used to generate a test evaluation report according to the data value evaluation report template and the evaluation result data.
[0099] In some implementations, generating a test evaluation report based on a data value evaluation report template and the evaluation result data includes: automatically filling the evaluation result data into the data value evaluation report template to generate a test evaluation report in a Word format.
[0100] The data evaluation system of Example 2 of the present invention can effectively ensure the security and efficiency of data collection, processing, transmission, and storage through automatic data extraction, encrypted transmission, evaluation result generation, and evaluation report generation; data extraction information is obtained through a visual interactive interface and sample customer data is automatically extracted, so that evaluators do not need to master advanced model development and algorithm skills, thereby reducing labor costs.
[0101] Through the description of the above implementation modes, those skilled in the art can clearly understand that the present invention can be implemented by combining software with a hardware platform. Based on such an understanding, all or part of the contribution of the technical solution of the present invention to the background technology can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation modes of the present invention or some parts of the implementation modes.
[0102] Correspondingly, an embodiment of the present invention further provides a computer-readable storage medium on which computer-readable instructions or programs are stored. When the computer-readable instructions or programs are executed by a processor, the computer performs the following operations: the operations include the steps included in the data evaluation method described in any of the above embodiments, which are not repeated here. The storage medium may include, for example, a CD, a hard disk, a floppy disk, a flash memory, a magnetic tape, etc.
[0103] In addition, an embodiment of the present invention further provides a computer device including a memory and a processor, wherein the memory is used to store one or more computer-readable instructions or programs, wherein the one or more computer-readable instructions or programs can implement the data evaluation method described in any of the above embodiments when executed by the processor. The computer device can be, for example, a server, a desktop computer, a notebook computer, a tablet computer, etc.
[0104] The present invention also provides a computer program product including a computer program, wherein the computer program includes program code for executing the data evaluation method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the data evaluation method provided by the embodiment of the present disclosure.
[0105] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).
[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention is described in detail with reference to the above embodiments, a person skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention. Therefore, the protection scope of the present invention shall be subject to the claims.
Claims
1. A data evaluation method, characterized in that: The data evaluation method comprises: Automatically calculate sample customer information in the business database according to a preset collection frequency, generate a sample customer information wide table based on the sample customer information, and automatically push the sample customer information wide table to the customer information library of the third-party data management platform for storage; Acquire data extraction information through a visual interactive interface, automatically extract sample data files from the customer information library according to the data extraction information, and store them in a test sample sampling library; The sample data file is automatically compressed and encrypted and then pushed to the data supplier operation end, and synchronously stored in the test sample sampling library; The data supplier operation terminal decrypts the compressed and encrypted sample data file, generates a sample feedback data file according to the decrypted sample data file, and stores the sample feedback data file in the test sample feedback library; Generate evaluation result data according to the data value evaluation algorithm template and the sample feedback data file; Generate a test evaluation report based on the data value evaluation report template and the evaluation result data.
2. The data evaluation method according to claim 1, characterized in that: Automatically extracting sample data files from the customer information database according to the data extraction information includes: encrypting the customer sensitive data in the sample data files and then extracting them.
3. The data evaluation method according to claim 1, characterized in that: Storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data file with the sample data file in the test sample sampling library according to the customer ID number identification and / or mobile phone identification of the sample customer.
4. The data evaluation method according to claim 1, characterized in that: Acquire the data value assessment algorithm template through an extensible data value assessment algorithm template library, wherein the extensible data value assessment algorithm template library includes: a value assessment algorithm template for scoring data, a value assessment algorithm template for variable data, and a value assessment algorithm template for blacklist data; The evaluation result data includes: sample data overview, coverage index, stability index, effectiveness index, relevance index, hit rate, accuracy rate, and recall rate index; Generating a test evaluation report according to the data value evaluation report template and the evaluation result data includes: automatically filling the evaluation result data into the data value evaluation report template to generate a test evaluation report.
5. A data evaluation system, characterized in that: The data evaluation system comprises: A sample customer information calculation module is used to automatically calculate sample customer information in the business database according to a preset collection frequency, generate a sample customer information wide table according to the sample customer information, and automatically push the sample customer information wide table to the customer information library of the third-party data management platform for storage; A sample data extraction module is used to obtain data extraction information through a visual interactive interface, automatically extract sample data files from the customer information library according to the data extraction information, and store them in a test sample sampling library; The sample data extraction module is also used to automatically compress and encrypt the sample data file and push it to the data supplier operation end, and synchronously store it in the test sample sampling library. The data supplier operation end decrypts the compressed and encrypted sample data file, generates a sample feedback data file according to the decrypted sample data file, and stores the sample feedback data file in the test sample feedback library; An evaluation result generation module, used to generate evaluation result data according to the data value evaluation algorithm template and the sample feedback data file; The evaluation report generation module is used to generate a test evaluation report based on the data value evaluation report template and the evaluation result data.
6. The data evaluation system according to claim 5, characterized in that Automatically extracting sample data files from the customer information database according to the data extraction information includes: encrypting the customer sensitive data in the sample data files and then extracting them.
7. The data evaluation system according to claim 5, characterized in that Storing the sample feedback data file in the test sample feedback library includes: automatically matching and storing the sample feedback data file with the sample data file in the test sample sampling library according to the customer ID number identification and / or mobile phone identification of the sample customer.
8. A computer-readable storage medium storing computer-readable instructions, characterized in that: The computer readable instructions are executed by a processor to implement the data evaluation method according to any one of claims 1 to 4.
9. A computer device comprising a memory and a processor, The memory stores computer-readable instructions, characterized in that: The processor executes the computer-readable instructions to implement the data evaluation method according to any one of claims 1 to 4.
10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the data evaluation method according to any one of claims 1 to 4 is implemented.
Citation Information
Cited By
Data source access test method, system and device of credit platform and medium
CN120856623A
