Method, device, equipment and medium for verifying annuity investment data

Abstract

The embodiments of the present application provide a method, device, equipment and medium for verifying annuity investment data, aiming to improve the validity of the annuity investment data stored in the database. The method comprises: upon receiving the target annuity investment data synchronized with the data source, respectively performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data; upon passing the data validity verification, obtaining information data corresponding to the target annuity investment data and historical position data corresponding to the target annuity investment data; performing a simulated valuation verification on the target annuity investment data based on the historical position data and the information data; wherein the simulated valuation verification is used to verify whether the target annuity investment data is data with normal valuation; upon passing the simulated valuation verification, storing the target annuity investment data in a preset database.

Description

Technical Field

[0001] The present application relates to the field of enterprise annuity technology, and in particular to a verification method, device, equipment and medium for annuity investment data. Background Art

[0002] Enterprise annuity is a supplementary pension system, which refers to the supplementary pension insurance system independently established by enterprises and their employees on the basis of participating in basic pension insurance in accordance with the law. When operating enterprise annuities, it is generally initiated by the principal and managed by the trustee under the supervision of the China Banking and Insurance Regulatory Commission. In the actual operation process, in addition to the principal and the trustee, it also involves the custodian, investment manager, account manager, etc.

[0003] Among them, the trustee can select, supervise and replace the account manager, custodian, investment manager and intermediary service agency, and regularly provide the enterprise annuity fund management report to the client, beneficiary and relevant regulatory authorities. Therefore, in order to better operate the enterprise annuity, the trustee and regulatory authorities generally need to provide service supervision on the work of annuity investment.

[0004] In the related art, the trustee will connect various remote digital investment data in order to conduct investment analysis. For example, the trustee will connect to the valuation data from the valuation system, the information data from the information provider, etc. Since the data collected by each data source vary greatly, the data synchronized from each data source are different, and effective statistical analysis of the annuity investment data cannot be performed. Therefore, there is an urgent need for an effective solution to verify the validity of the annuity investment data from each data source. Summary of the invention

[0005] In order to solve the above problems, the present application provides a method, device, equipment and medium for verifying annuity investment data, which is intended to verify the validity of annuity investment data from various data sources.

[0006] In a first aspect of an embodiment of the present application, a method for verifying annuity investment data is provided, the method comprising:

[0007] When receiving the target annuity investment data synchronized with the data source, respectively performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data;

[0008] When the data validity check passes, obtaining information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data;

[0009] Based on the historical holding data and the information data, a simulated valuation check is performed on the target annuity investment data; wherein the simulated valuation check is used to check whether the target annuity investment data is data with normal valuation;

[0010] When the simulation valuation verification passes, the target annuity investment data is stored in a preset database.

[0011] Optionally, performing data validity verification on the data belonging to preset dimensions of the target annuity investment data respectively includes:

[0012] Extracting the account name in the target annuity investment data, and verifying the validity of the account name in the target annuity investment data;

[0013] When the validity check of the account name passes, the validity of the data format and / or value of each preset field in the target annuity investment data is checked.

[0014] Optionally, verifying the validity of the account name of the target annuity investment data includes:

[0015] Determining whether the account name belongs to a plurality of preset account names in a preset rule base, and / or obtaining a text vector of the account name, and inputting the text vector into a named entity recognition model to verify the validity of the account name;

[0016] Verifying the validity of the data format and / or value of each preset field in the target annuity investment data includes:

[0017] Read the historical investment data corresponding to the account name;

[0018] Based on the field value of the preset field in the historical investment data, the field value of the preset field in the target annuity investment data is verified.

[0019] Optionally, based on the field value of the preset field in the historical investment data, verifying the field value of the preset field in the target annuity investment data includes:

[0020] Inputting the historical investment data and the target annuity investment data into a data anomaly detection model to perform data anomaly verification on the target annuity investment data;

[0021] The data anomaly detection model is obtained by training a first preset model using a plurality of historical abnormal investment data as training samples.

[0022] Optionally, the target annuity investment data includes target transaction data and target valuation data; based on the historical holding data and the information data, a simulation valuation check is performed on the target annuity investment data, including:

[0023] Based on the information data, the historical position data is updated;

[0024] Based on the updated historical position data and the target transaction data, the valuation data is extrapolated to obtain estimated valuation data;

[0025] Based on the estimated valuation data, the validity of the target valuation data is verified.

[0026] Optionally, based on the valuation data, verifying the validity of the target annuity investment data includes:

[0027] Comparing the estimated valuation data with the target valuation data to obtain a difference;

[0028] Based on the difference and a preset difference threshold, a verification result of the validity of the target valuation data is determined.

[0029] Optionally, the valuation data is extrapolated based on the updated historical position data and the target transaction data to obtain estimated valuation data, including:

[0030] Preprocessing the information data and the updated historical position data respectively to obtain standardized information data and standardized historical position data;

[0031] Inputting the standardized information data and the historical position data into a position data prediction model to obtain estimated valuation data output by the position data prediction model;

[0032] The position data prediction model is obtained by training a second preset model with multiple historical position data samples and corresponding multiple information data samples.

[0033] According to a second aspect of the embodiments of the present application, a verification device for annuity investment data is provided, the device comprising:

[0034] A first verification module is used to perform data validity verification on data belonging to a preset dimension of the target annuity investment data respectively when receiving the target annuity investment data synchronized with the data source;

[0035] A data acquisition module, used to acquire information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data when the data validity check passes;

[0036] A second verification module is used to perform a simulated valuation verification on the target annuity investment data based on the historical holding data and the information data; wherein the simulated valuation verification is used to verify whether the target annuity investment data is data with normal valuation;

[0037] The storage module is used to store the target annuity investment data into a preset database when the simulation valuation verification passes.

[0038] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when executed, the processor implements the method for verifying annuity investment data as described in the first aspect.

[0039] According to a fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided, wherein a computer program stored in the medium enables a processor to execute the verification method for annuity investment data as described in the first aspect.

[0040] The technical solution of the embodiment of the present application has at least the following advantages:

[0041] In an embodiment of the present application, when the target annuity investment data synchronized with the data source is received, data validity verification can be performed on the data belonging to the preset dimensions of the target annuity investment data respectively; when the data validity verification passes, information data corresponding to the target annuity investment data and historical position data are obtained; then, based on the historical position data and information data, a data simulation valuation verification is performed on the target annuity investment data; wherein, the simulation valuation verification is used to verify whether the target annuity investment data is data with normal valuation; when the simulation valuation verification passes, the target annuity investment data is stored in a preset database.

[0042] After receiving the synchronized annuity investment data, the data validity of the preset dimensions in the target annuity investment data is first checked to determine whether the target annuity investment data is valid data, wherein the preset dimensions may refer to the account name, data format, etc. of the annuity investment data. After determining that the target annuity investment data is valid data, the annuity investment data may be subjected to a simulated valuation check based on historical position data and information data, so as to determine whether the target annuity investment data to be stored currently is data with normal valuation through the historical position data and information data that have passed the verification. If it is data with normal valuation, the target annuity investment data is stored, thereby ensuring the subsequent effective statistical analysis of the annuity investment plan based on the annuity investment data. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0044] Figure 1 is a schematic diagram of a communication environment used by a verification method for annuity investment data according to an embodiment of the present application;

[0045] Figure 2 This is a schematic diagram of the steps of a method for verifying annuity investment data according to an embodiment of the present application;

[0046] Figure 3 It is a schematic diagram of the overall process of a verification method of annuity investment data shown in one embodiment of the present application;

[0047] Figure 4 is a flowchart of the steps of performing simulation valuation verification shown in an embodiment of the present application;

[0048] Figure 5 It is a schematic diagram of the framework of a verification device for annuity investment data shown in one embodiment of the present application. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims.

[0051] In the related art, the annuity investment data from multiple data sources vary greatly. Generally, it is necessary to parse and format these annuity investment data before statistical analysis of the annuity data. However, after parsing and formatting, it is often found that the annuity investment data obtained after processing is erroneous, that is, it cannot be used for statistical analysis. In this case, the only way is to rely on staff to trace back the data to eliminate the errors. However, this method is very time-consuming and labor-intensive, and inefficient.

[0052] In view of this, the applicant proposes a solution for verifying the validity of annuity investment data before parsing and formatting the annuity investment data, so as to proactively verify the data validity and simulated valuation of the annuity investment data before the annuity investment data is entered into the database, thereby ensuring that the annuity investment data stored in the database is valid and usable, thereby reducing the probability of errors and improving the efficiency of investment statistical analysis.

[0053] Reference Figure 1 As shown, a schematic diagram of the communication environment used by the verification method of annuity investment data provided by the present application is shown, such as Figure 1 As shown, it includes multiple different data sources, and an investment supervision and risk management platform that connects to multiple different data sources, wherein the investment supervision and risk management platform is managed and maintained by the trustee in the enterprise annuity, wherein the multiple different data sources include: valuation system, on-site investment trading system, alternative investment trading system, credit rating system, information system and other external data sources.

[0054] In practice, when synchronizing the annuity investment data of various data sources, the investment supervision and risk management platform can obtain it by scheduled pulling or scheduled push of data sources, or by web crawlers. The annuity investment data of the synchronized data sources can be data of various file types, for example, it can include txt, excel, Html and other file types, which are not limited here.

[0055] Combination Figure 1 , the annuity investment data verification method of the present application embodiment is introduced, referring to Figure 2 As shown, a flowchart of the steps of the verification method of annuity investment data is shown, such as Figure 2 As shown, the following steps may be specifically included:

[0056] Step S201: upon receiving the target annuity investment data synchronized with the data source, respectively performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data.

[0057] In this embodiment, the data source can be Figure 1Among the multiple data sources shown, the target annuity investment data may refer to the investment data of the target annuity plan, which may include valuation data and position data transmitted by the valuation system.

[0058] The preset dimension may refer to a dimension in the target annuity investment data related to data validity, such as an account name dimension, a numerical format dimension, etc. In specific implementation, it may be detected whether the data in the target annuity investment data belonging to the preset dimension is valid, such as whether the account name is valid or whether the numerical format is correct.

[0059] Step S202: When the data validity check passes, obtain information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data.

[0060] In this embodiment, when the data belonging to the preset dimension in the target annuity investment data passes the data validity check, it indicates that the data in the target annuity investment data is valid, for example, the account name is valid and the numerical format is correct. In this way, information data and historical position data can be further obtained. Among them, the information data and historical position data and the target annuity investment data are all data belonging to the same annuity plan of the same account, and the information data and historical position data are all data that have passed the check and have been stored.

[0061] Among them, the information data can be the market change information of each asset corresponding to the investment portfolio in the annuity plan, and can reflect the equity change details of each asset. The information data includes information such as transaction price, closing price, equity change, etc., and can reflect the market change information of each asset, such as the change information of stocks launched by Company A and stocks launched by Company B in the market. For example, if the asset is a stock, it can include details such as the equity registration date and the ex-rights and ex-dividend date.

[0062] Historical position data refers to the historical change data of the position of the annuity plan, for example, it can be the position data of the previous trading day. Among them, the position held by the client is due to buying and selling, and the positions after these buying and selling constitute the position data. Therefore, the position data can reflect the position details of multiple assets held by the client, for example, including the number of positions, holding costs, and holding market value of each asset. Among them, assets can be understood as actual securities categories such as stocks, bonds, and futures. Generally speaking, position data is counted in time periods. For example, the position is counted once a day, so that there is a position data every day. By counting the positions of multiple days, the changes in the positions over multiple days can be obtained.

[0063] Step S203: Based on the historical holding data and the information data, a simulation valuation check is performed on the target annuity investment data.

[0064] The simulated valuation verification is used to verify whether the target annuity investment data is normally valued data.

[0065] In this embodiment, after obtaining the information data and historical position data, the information data and historical position data, as well as the position data in the target annuity investment data, can be used to predict the latest transaction data received, and thus, based on the latest transaction data received, the valuation position data that should be received this time is predicted, and compared with the valuation position data actually received in the target annuity investment data to perform a simulated valuation verification. If the deviation is large, it indicates that the target annuity investment data is not normal valuation data, and an alarm message is given, and the simulated valuation verification fails. If the deviation is small, it indicates that the target annuity investment data is normal valuation data, and the simulated valuation verification passes.

[0066] Step S204: When the simulation valuation verification passes, the target annuity investment data is stored in a preset database.

[0067] In this embodiment, when the simulation valuation verification passes, the data representing the target annuity investment data in the preset dimensions is valid and normal valuation data. For example, the account name and the numerical format are both valid and compared with the historical investment data, it is also normal valuation data. Therefore, it can be used for subsequent investment statistical analysis, so that the target annuity investment data can be stored in a preset database for subsequent investment statistical analysis.

[0068] According to the technical solution of the present embodiment, after receiving the synchronized annuity investment data, the data validity of the preset dimensions in the target annuity investment data is first checked to determine whether the target annuity investment data is valid data, wherein the preset dimensions may refer to the account name, data format, etc. of the annuity investment data. After determining that the target annuity investment data is valid data, the annuity investment data may be subjected to a simulated valuation check based on historical position data and information data, so as to determine whether the target annuity investment data to be stored currently is data with normal valuation through the historical position data and information data that have passed the verification. If it is data with normal valuation, the target annuity investment data is stored, thereby ensuring the subsequent effective statistical analysis of the annuity investment plan based on the annuity investment data.

[0069] Reference Figure 3 As shown, Figure 3 FIG. 1 shows a schematic diagram of the overall process of a method for verifying annuity investment data in an embodiment. Figure 3As shown, after synchronization to the target annuity investment data, the account validity of the target annuity investment data can be verified through the named entity recognition module. After the verification is passed, the format and value of the data in the target annuity investment data can be verified for validity through the data anomaly detection module. After the verification is passed, the valuation of the target annuity investment data can be verified whether it is normal through the simulation valuation module. After the verification is passed, investment analysis and statistics can be performed on the target annuity investment data.

[0070] When any verification link fails, an alarm message may be output to remind the user that the verification fails.

[0071] Combination Figure 3 As shown, in one embodiment, the process of performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data may be:

[0072] First, the account name in the target annuity investment data is extracted, and the validity of the account name in the target annuity investment data is verified.

[0073] In this embodiment, the preset dimension may refer to a dimension in the target annuity investment data related to data validity, such as an account name dimension, a numerical format dimension, etc. In practice, the account name in the target annuity investment data may be extracted, and the validity of the account name may be verified. The account name may be an account of the principal to whom the target annuity investment data belongs, which may refer to an account opened in a bank or an account opened in a trustee.

[0074] In practice, the account name can be identified from the target annuity investment data, and then the validity of the account name can be verified. Specifically, it can be determined whether the identified account name is in the account library, which can store multiple account names opened by the client. If it is in the account library, it means that the account name is valid, that is, there is no problem with the account of the target annuity investment data.

[0075] Afterwards, when the validity check of the account name is passed, the validity of the data format and / or value of each preset field in the target annuity investment data is checked.

[0076] Among them, the preset fields can be set according to actual conditions, such as the preset fields include yield, transaction amount, expenditure amount, etc.; among them, the data format of the preset field can refer to the data format of the field value of the preset field, and the value of the preset field can refer to the size of the field value of the preset field.

[0077] Among them, the format of the numerical value in the target annuity investment data can be determined. Specifically, it can be the format of the numerical value related to the amount. If the format of the numerical value related to the amount is numeric, such as floating point, it is normal and valid data. If not, it is abnormal and invalid data.

[0078] In practice, the formats of the various numerical values ​​in the target annuity investment data can also be combined with the formats of the various numerical values ​​in the historical annuity investment data that have been verified to perform consistency verification. For example, the transaction amount in the historical annuity investment data stored in the database has always been a floating point, while the transaction amount in the target annuity investment data is a string, then it can be determined that the transaction amount in the target annuity investment data is abnormal and the verification fails.

[0079] In this embodiment, when the account name of the target annuity investment data is a valid account and the format of the value is correct, the data validity verification of the target annuity investment data is passed. In other cases, if the account name is valid but the format of the value is not valid, the data validity verification of the target annuity investment data may fail. When the data validity verification fails, an alarm message may be sent, and the result of the data validity verification failure may be displayed so that the data verification can be manually performed.

[0080] Among them, after the manual data verification is passed, the next step of simulation valuation verification can be carried out.

[0081] In a specific implementation method, when verifying the validity of the account name of the target annuity investment data, it can be determined whether the account name belongs to multiple preset account names in a preset rule library, and / or, a text vector of the account name is obtained, and the text vector is input into a named entity recognition model to verify the validity of the account name.

[0082] In this implementation, the validity check of the account name can be performed in the following three ways:

[0083] The first method is to determine whether the account name belongs to multiple preset account names in the preset rule base. The preset rule base includes multiple legal account names opened by the client. If the account name of the target annuity investment data is in the preset rule base, the validity check of the account name passes.

[0084] The second way is to obtain a text vector of the account name and input the text vector into a named entity recognition model to verify the validity of the account name.

[0085] In practice, since the preset rule base cannot exhaust all the legal account names opened by the client, for example, when the account name of the target annuity investment data is a new account opened by the client, the account name will not exist in the preset rule base. For another example, when the account name of the target annuity investment data is a full name, but the preset rule base uses an abbreviation, there will be a false check. In practice, a neural network model can be used to check the validity of the account name.

[0086] Specifically, the account name can be extracted from the target annuity investment data and the account name can be text-vectorized, for example, the account name can be converted into a word vector or a sentence vector to obtain a text vector. After that, the text vector is input into a named entity recognition model, and the named entity recognition model is used to verify the legitimacy of the text vector to determine whether the account name corresponding to the text vector is valid and legal.

[0087] The process of training the named entity recognition model may be:

[0088] Collect multiple account name samples, each of which carries a label that can characterize whether the account name sample is a legal account name sample or an illegal account name sample. Then, convert the multiple account name samples into text vector samples, and then use the text vector samples to train the preset neural network model. During each training, the loss function can be calculated based on the output of the neural network model and the label of the account name sample, and then the parameters in the neural network model are updated based on the calculated loss value, thereby obtaining a named body recognition model.

[0089] In the third method, the validity check of the account name is first performed using the first method, and when the first method fails to check, the validity check of the account name is performed using the second method. That is, first determine whether the account name belongs to multiple preset account names in the preset rule library, and if not, obtain the text vector of the account name, and input the text vector into the name recognition model to check the validity of the account name.

[0090] When the third method is adopted, the efficiency of account name verification can be improved by combining the preset rule base and the named entity recognition model, which can greatly avoid manual re-verification.

[0091] Of course, when the third method still shows that the data validity check fails, an alarm message can be sent, and the result of the data validity check failure can be displayed so that the data can be manually checked. When the result of the manual check is obtained, the result of the manual check can be used as the label of the account name of the target annuity investment data, so that the account name of the target annuity investment data is used as a training sample to update the named body recognition model, so that the named body recognition model can recognize the validity of the account name in this case, thereby enhancing the generalization performance of the named body recognition model.

[0092] In a specific implementation method, when verifying the data format and / or validity of the numerical value of each preset field in the target annuity investment data, the historical investment data corresponding to the account name can be read; and based on the field value of the preset field name in the historical investment data, the field value of the preset field name in the target annuity investment data can be verified.

[0093] In this embodiment, as described above, the validity of the data format and / or value of the preset field in the target annuity investment data can be verified based on the historical investment data that has been verified in the historical process. In specific implementation, it can be determined whether the data format of the preset field in the historical investment data is consistent with the data format of the preset field in the target annuity investment data, and whether the difference between the field value of the preset field in the historical investment data and the field value of the preset field in the target annuity investment data is less than the preset difference.

[0094] Among them, when the data format of the preset fields in the historical investment data is consistent with the data format of the preset fields in the target annuity investment data, and the difference between the field values ​​of the preset fields is less than the preset difference, it is determined that the validity check of the data format and / or value of each preset field in the target annuity investment data has passed.

[0095] Among them, the situation where the difference between the field values ​​of the preset fields is less than the preset difference indicates that the numerical difference between the same item in the historical investment data and the target annuity investment data is small. For example, taking the net asset value as an example, if the net asset value changes from 1000 in the historical investment data to negative 10000 in the target annuity investment data, it is considered that there is a serious error in the net asset value of the target annuity investment data, which is invalid data and can fail the verification.

[0096] In one embodiment, based on the field value of the preset field in the historical investment data, the implementation process of verifying the field value of the preset field in the target annuity investment data may be as follows:

[0097] The historical investment data and the target annuity investment data are input into a data anomaly detection model to perform data anomaly verification on the target annuity investment data.

[0098] The data anomaly detection model is obtained by training a first preset model using a plurality of historical abnormal investment data as training samples.

[0099] In this embodiment, the data anomaly detection model can be used to verify the validity of the preset fields in the target annuity investment data. The data anomaly detection model can be trained in the following way:

[0100] Collect multiple investment data sample pairs, wherein each investment data sample pair includes investment data to be verified and investment data as a reference, and each investment data sample pair carries a label, which can indicate whether a preset field in the investment data to be verified is legal or illegal. After that, process the multiple investment data sample pairs into samples that can be processed by a neural network, for example, convert them into a preset data structure, and then use the investment data sample pairs to train a first preset model, which can be used to extract the preset fields of the two investment data samples in the investment data sample pair, and then compare the two extracted preset fields, so as to verify whether the preset fields in the investment data to be verified are legal.

[0101] During each training, the loss function can be calculated based on the output of the first preset model and the label of the investment data sample pair, and then the parameters in the first preset model are updated according to the calculated loss value, so as to obtain a data anomaly detection model.

[0102] In the present application, after obtaining the data anomaly detection model, the historical investment data and the target annuity investment data can be input into the data anomaly detection model to perform data anomaly verification on the data format and / or field value size of the preset fields in the target annuity investment data.

[0103] In practice, a combination of predefined rules and data anomaly detection models can be used to verify the validity of preset fields in the target annuity investment data. Specifically, the field values ​​of the preset fields in the target annuity investment data can be verified based on the field values ​​of the preset fields in the historical investment data. If the verification fails, the historical investment data and the target annuity investment data are input into the data anomaly detection model to perform data anomaly verification on the target annuity investment data through the data anomaly detection model.

[0104] When this implementation is adopted, the efficiency of the preset field validity check can be improved by utilizing a combination of field value verification and data anomaly detection model, thereby greatly avoiding manual re-verification. Of course, when the preset field validity check still fails, an alarm message can be sent, and the result of the preset field validity check failure can be displayed to facilitate manual data verification. When the result of manual verification is obtained, whether it is a result of passing the verification or a different result, the target annuity investment data and the historical investment data can be used as a training sample pair, and the result of the manual verification can be used as a label for the sample pair, thereby updating the data anomaly detection model so that the data anomaly detection model can identify the validity of the preset field in this case, thereby enhancing the generalization performance of the data anomaly detection model.

[0105] When this implementation is adopted, the validity of the preset fields in the target annuity investment data can be verified based on the historical investment data, thereby improving the verification accuracy of the target annuity investment data, and detecting anomalies in data format and numerical fluctuations, thereby ensuring the validity and availability of the stored investment data.

[0106] Next, we will introduce how to perform simulation valuation verification. Figure 4 As shown, a flowchart of the steps of performing simulation valuation verification is shown, wherein the target annuity investment data includes target transaction data and target valuation data; Figure 4 As shown, the following steps may be specifically included:

[0107] Step S401: Based on the information data, the historical position data is updated.

[0108] In this embodiment, since the information data is the market change information of each asset corresponding to the investment portfolio in the annuity plan, it can reflect the equity change details of each asset. Therefore, the historical holding data that has been stored can be updated based on the consulting data.

[0109] For example, when updating historical position data based on information data, it is necessary to update the corresponding positions according to different asset categories, where:

[0110] For stocks: you can use the stock code of the position to associate with the code in the information data to obtain the stock quotes in the information data, and then update the market value in the historical position data. The market value in the historical position data = quantity * price in the stock quotes.

[0111] For interest-bearing bonds: you can use the market code of the held bonds to associate the basic information of Caihui bonds in the information data to get the bond interest rate, and then update the interest calculation of the interest-bearing bonds in the historical holdings data. Interest calculation = quantity * face value * bond interest rate / number of days in a year.

[0112] For deposits, claims and trusts: you can use the valuation securities information table in the internal code-related information data of held claims and trusts to obtain interest rate information, and then update the interest calculation of deposits, claims and trusts in the historical holdings data. Interest calculation = principal * interest rate / 365.

[0113] For convertible bonds: you can use the market price table code in the information data associated with the stock holding code to obtain the market price, and then update the market value of the convertible bonds in the historical holding data. Market value = quantity * price.

[0114] Fees: Calculated using combined rates.

[0115] In this way, updated historical holdings data can be obtained, wherein the updated historical holdings data can reflect the latest annuity plan holdings data obtained based on information data under the current circumstances. The latest annuity plan holdings data is estimated and can be used as a reference for the target annuity investment data.

[0116] Step S402: Based on the updated historical position data and the target transaction data, the valuation data is extrapolated to obtain estimated valuation data.

[0117] In this embodiment, the target annuity investment data includes target transaction data and target valuation data, wherein the target transaction data and the target valuation data are both real investment data to be stored. The updated historical holdings data can reflect the latest holdings data of the annuity plan obtained based on the information data under the current circumstances. Therefore, the valuation data can be further calculated based on the updated historical holdings data and the target transaction data in the target annuity investment data, thereby obtaining the estimated valuation data.

[0118] Since the valuation data is the data for valuing individual investment assets in the annuity plan, and the position data reflects the asset changes before and after the investment assets are traded, and the transaction data records the data of the transaction changes of each investment asset, the target transaction data is combined with the position data to calculate the valuation data; the estimated valuation data is determined based on the real and valid position data that has been stored, the real and valid information data and the target transaction data to be stored, and can reflect the reasonable range of the valuation data to be stored.

[0119] Step S403: Based on the estimated valuation data, verify the validity of the target valuation data.

[0120] In this embodiment, since the estimated valuation data can reflect the reasonable range of the valuation data to be stored, the estimated valuation data can be used as a reference to compare it with the target valuation data. Specifically, the difference between the two can be compared. When the difference is less than the preset difference threshold, the validity check of the target valuation data is passed. When the difference is large, it indicates that the target valuation data is not within the estimated reasonable range and may be erroneous and abnormal valuation data.

[0121] Accordingly, in one embodiment, when verifying the validity of the target annuity investment data based on the valuation data, the estimated valuation data and the target valuation data can be compared to obtain the difference; and based on the difference and a preset difference threshold, the verification result of the validity of the target valuation data can be determined.

[0122] The difference may refer to the difference between the field values ​​of the same field in the estimated valuation data and the target valuation data. The preset difference threshold may be set according to the actual situation. Since the simulation valuation check is to check the obviously unreasonable valuation data, the preset difference threshold may be set larger.

[0123] When this implementation is adopted, the target valuation data in the target investment data to be stored can be verified for validity based on the historical position data and current information data that have been stored to verify whether it is reasonable. In this way, the accuracy and comprehensiveness of the verification of the investment data can be improved.

[0124] Correspondingly, in one embodiment, when the valuation data is extrapolated based on the information data, the updated historical position data and the target transaction data to obtain the estimated valuation data, the information data and the updated historical position data can be preprocessed respectively to obtain standardized information data and standardized historical position data; and the standardized information data and the historical position data can be input into a position data prediction model to obtain the estimated valuation data output by the position data prediction model.

[0125] The position data prediction model is obtained by training a second preset model with multiple historical position data samples and corresponding multiple information data samples.

[0126] In this embodiment, preprocessing the information data may refer to converting the information data into a preset data structure, for example, converting the information data into an Excel data structure, and preprocessing the updated historical position data may refer to converting the updated historical position data into a preset data structure, for example, an Excel data structure, so that the data input into the position data prediction model is data of a standard data structure.

[0127] In the embodiment of the present application, the process of training and obtaining the position data prediction model can be as follows:

[0128] Collect multiple sets of training samples, each set of training samples includes historical position data samples and information data samples and corresponding valuation data samples, then use the valuation data samples as labels, input the historical position data samples and information data samples in each set of training samples into the second preset model, the second preset model is used to process the historical position data samples and information data samples according to the preset algorithm to output the estimated valuation data, then calculate the loss function according to the output estimated valuation data and valuation data samples, then update the parameters of the second preset model according to the calculated loss value, and the second preset model after multiple updates is the position data prediction model.

[0129] Among them, the position data prediction model needs to be trained regularly with historical actual position data and actual information data to adapt to the impact of changes in market and internal product positions.

[0130] When the technical solution of this embodiment is adopted, the neural network model can be used to predict the estimated valuation data, thereby improving the generalization ability and computing efficiency.

[0131] The technical solution of the embodiment of the present application has the following advantages:

[0132] 1. Use the technology combining neural network and predefined rules to verify the target annuity investment data at multiple levels, so as to find and solve the quality problems of investment transaction data as early as possible. In addition, the use of predefined rules (preset rule base, etc.) can realize the rapid detection of data quality; neural network can be used as a supplement when the predefined rules fail, and through the continuous correction and improvement of the historical data stored in the library, the performance of data quality detection is effectively improved.

[0133] 2. It can effectively improve the accuracy of statistical analysis of investment data and greatly reduce the workload of backtracking when problems arise in data statistical reports, thereby improving the accuracy and timeliness of investment decisions and risk management.

[0134] Based on the same inventive concept, in one embodiment, a verification device for annuity investment data is also provided, referring to Figure 5 FIG. 1 is a block diagram showing a structure of a verification device for annuity investment data, the device comprising:

[0135] The first verification module 501 is used to perform data validity verification on data belonging to a preset dimension of the target annuity investment data respectively when receiving the target annuity investment data synchronized with the data source;

[0136] A data acquisition module 502, for acquiring information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data when the data validity check passes;

[0137] The second verification module 503 is used to perform a simulation valuation verification on the target annuity investment data based on the historical holding data and the information data; wherein the simulation valuation verification is used to verify whether the target annuity investment data is data with normal valuation;

[0138] The storage module 504 is used to store the target annuity investment data into a preset database when the simulation valuation verification passes.

[0139] Optionally, the first verification module 501 includes:

[0140] a first verification unit, configured to extract an account name from the target annuity investment data and verify the validity of the account name of the target annuity investment data;

[0141] The second verification unit is used to verify the validity of the data format and / or value of each preset field in the target annuity investment data when the validity verification of the account name passes.

[0142] Optionally, the first verification unit is specifically used to determine whether the account name belongs to a plurality of preset account names in a preset rule base, and / or obtain a text vector of the account name and input the text vector into a named entity recognition model to verify the validity of the account name;

[0143] The second verification unit is specifically used to read the historical investment data corresponding to the account name; and based on the field value of the preset field in the historical investment data, verify the field value of the preset field in the target annuity investment data.

[0144] Optionally, the second verification unit is specifically used to input the historical investment data and the target annuity investment data into a data anomaly detection model to perform data anomaly verification on the target annuity investment data;

[0145] The data anomaly detection model is obtained by training a first preset model using a plurality of historical abnormal investment data as training samples.

[0146] Optionally, the target annuity investment data includes target transaction data and target valuation data; the second verification module 503 includes:

[0147] An updating unit, configured to update the historical position data based on the information data;

[0148] An estimation unit, configured to calculate the valuation data based on the updated historical position data and the target transaction data to obtain estimated valuation data;

[0149] A verification unit is used to verify the validity of the target valuation data based on the estimated valuation data.

[0150] Optionally, the verification unit is specifically used to compare the estimated valuation data with the target valuation data to obtain a difference; and based on the difference and a preset difference threshold, determine a verification result of the validity of the target valuation data.

[0151] Optionally, the updating unit includes:

[0152] A preprocessing subunit, used to preprocess the information data and the updated historical position data respectively to obtain standardized information data and standardized historical position data;

[0153] An input subunit, used to input the standardized information data and the historical position data into a position data prediction model to obtain estimated valuation data output by the position data prediction model;

[0154] The position data prediction model is obtained by training a second preset model with multiple historical position data samples and corresponding multiple information data samples.

[0155] The annuity investment data verification device of the embodiment of the present application, after receiving the synchronized annuity investment data, first performs a data validity verification on the data of the preset dimensions in the target annuity investment data to determine whether the target annuity investment data is valid data, wherein the preset dimensions may refer to the account name, data format, etc. of the annuity investment data. After determining that the target annuity investment data is valid data, a simulated valuation verification can be performed on the annuity investment data based on historical position data and information data, so as to determine whether the target annuity investment data to be stored currently is data with normal valuation through the historical position data and information data that have passed the verification. If it is data with normal valuation, the target annuity investment data is stored again, thereby ensuring the subsequent effective statistical analysis of the annuity investment plan based on the annuity investment data.

[0156] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0157] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0158] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0159] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0160] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0161] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0162] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0163] The verification method, device, equipment and medium for annuity investment data provided by the present invention are introduced in detail above. Specific examples are used herein to illustrate the principle and implementation mode of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation mode and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for verifying annuity investment data, characterized in that: The method comprises: When receiving the target annuity investment data synchronized with the data source, respectively performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data, the target annuity investment data including the target transaction data and the target valuation data; When the data validity check passes, obtaining information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data; Based on the historical holding data and the information data, a simulated valuation check is performed on the target annuity investment data; wherein the simulated valuation check is used to check whether the target annuity investment data is data with normal valuation; When the simulation valuation verification passes, storing the target annuity investment data in a preset database; The performing data validity verification on the data belonging to the preset dimensions of the target annuity investment data respectively includes: Extracting the account name in the target annuity investment data, and verifying the validity of the account name in the target annuity investment data; When the validity check of the account name passes, verifying the validity of the data format and / or value of each preset field in the target annuity investment data; The verifying the validity of the account name of the target annuity investment data includes: Determining whether the account name belongs to a plurality of preset account names in a preset rule base, and / or obtaining a text vector of the account name, and inputting the text vector into a named entity recognition model to verify the validity of the account name; The checking of the validity of the data format and / or the value of each preset field in the target annuity investment data includes: Read the historical investment data corresponding to the account name; Based on the field value of the preset field in the historical investment data, verifying the field value of the preset field in the target annuity investment data; The simulation valuation verification of the target annuity investment data based on the historical holding data and the information data includes: Based on the information data, the historical position data is updated; Based on the updated historical position data and the target transaction data, the valuation data is extrapolated to obtain estimated valuation data; Based on the estimated valuation data, the validity of the target valuation data is verified.

2. The method according to claim 1, characterized in that: Based on the field value of the preset field in the historical investment data, verifying the field value of the preset field in the target annuity investment data includes: Inputting the historical investment data and the target annuity investment data into a data anomaly detection model to perform data anomaly verification on the target annuity investment data; The data anomaly detection model is obtained by training a first preset model using a plurality of historical abnormal investment data as training samples.

3. The method according to claim 1, characterized in that Based on the estimated valuation data, verifying the validity of the target valuation data includes: Comparing the estimated valuation data with the target valuation data to obtain a difference; Based on the difference and a preset difference threshold, a verification result of the validity of the target valuation data is determined.

4. The method according to claim 1, characterized in that: Based on the updated historical position data and the target transaction data, the valuation data is extrapolated to obtain estimated valuation data, including: Preprocessing the information data and the updated historical position data respectively to obtain standardized information data and standardized historical position data; Inputting the standardized information data and the historical position data into a position data prediction model to obtain estimated valuation data output by the position data prediction model; The position data prediction model is obtained by training a second preset model with multiple historical position data samples and corresponding multiple information data samples.

5. A verification device for annuity investment data, characterized in that: The device comprises: A first verification module is used to perform data validity verification on data belonging to preset dimensions of the target annuity investment data respectively when receiving the target annuity investment data synchronized with the data source, wherein the target annuity investment data includes target transaction data and target valuation data; A data acquisition module, used to acquire information data corresponding to the target annuity investment data and historical holding data corresponding to the target annuity investment data when the data validity check passes; A second verification module is used to perform a simulated valuation verification on the target annuity investment data based on the historical holding data and the information data; wherein the simulated valuation verification is used to verify whether the target annuity investment data is data with normal valuation; A storage module, used for storing the target annuity investment data in a preset database when the simulation valuation verification passes; The first verification module includes: a first verification unit, configured to extract an account name from the target annuity investment data and verify the validity of the account name of the target annuity investment data; A second verification unit, configured to verify the validity of the data format and / or value of each preset field in the target annuity investment data when the validity verification of the account name passes; The first verification unit is specifically used to determine whether the account name belongs to a plurality of preset account names in a preset rule base, and / or obtain a text vector of the account name and input the text vector into a named entity recognition model to verify the validity of the account name; The second verification unit is specifically configured to read the historical investment data corresponding to the account name; and verify the field value of the preset field in the target annuity investment data based on the field value of the preset field in the historical investment data; The second verification module includes: An updating unit, configured to update the historical position data based on the information data; An estimation unit, configured to calculate the valuation data based on the updated historical position data and the target transaction data to obtain estimated valuation data; A verification unit is used to verify the validity of the target valuation data based on the estimated valuation data.

6. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor is executed, the verification method for annuity investment data as described in any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium, characterized in that: The computer program stored therein enables the processor to execute the verification method for annuity investment data as described in any one of claims 1-4.

Images