Enterprise financial sharing evaluation method and system based on big data

Abstract

The invention relates to the technical field of data sharing evaluation, and discloses an enterprise financial sharing evaluation method and system based on big data, and the method comprises the steps: analyzing historical enterprise financial data based on a decision tree algorithm, and building a data quality measurement standard exclusive to a specific data source and a specific data type; according to the method, a data quality measurement standard is obtained, and a dynamic data quality evaluation mechanism is constructed based on the data quality measurement standard, so that enterprise financial data needing to be shared are screened through two layers of mechanisms, the data quality and the data quality evaluation accuracy of enterprise financial data sharing evaluation are ensured, and the accuracy of enterprise financial data sharing evaluation is further improved; in addition, access control and encryption operation are performed on the data, so that the security and privacy of the data are further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data sharing evaluation, and in particular to a method and system for enterprise financial sharing evaluation based on big data. Background Art

[0002] Financial Data Sharing Assessment refers to the systematic evaluation and analysis of the financial data sharing process between internal departments or external partners of an enterprise to ensure the effectiveness, security and compliance of data sharing.

[0003] In the evaluation of enterprise financial data sharing, data quality assessment is one of the most important links. The purpose of data quality assessment is to ensure the accuracy, completeness and consistency of the shared data, and to ensure that the shared data is up-to-date, correct, and consistent between different systems.

[0004] At present, in the process of enterprise financial data sharing and evaluation, due to the large number of data sources, data inconsistency or data duplication is prone to occur during the data integration process, affecting the accuracy of subsequent data analysis results. When dealing with large-scale, rapidly changing financial data, there are still challenges in processing and evaluating real-time data. In addition, existing data quality assessments are usually static, and lack real-time monitoring and evaluation of dynamic changes in data, making it impossible to detect and correct data quality problems in a timely manner. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide an enterprise financial sharing evaluation method and system based on big data, so as to solve the problems that data quality evaluation cannot be dynamically evaluated and the evaluation accuracy is not high in the current enterprise financial data sharing evaluation process.

[0006] The first aspect of the present invention discloses a method for evaluating enterprise financial sharing based on big data, the method comprising the following steps:

[0007] Step S1: establishing a data quality metric standard, and screening the enterprise financial data to be shared based on the data quality metric standard to obtain first data;

[0008] Step S2, performing automated cleaning and preprocessing operations on the first data to obtain second data;

[0009] Step S3: establishing a dynamic data quality evaluation mechanism based on the data quality metric standard, and performing a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data;

[0010] Step S4, setting access control on the third data to obtain fourth data;

[0011] Step S5: perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

[0012] Furthermore, the operation of establishing a data quality metric standard in step S1 specifically includes analyzing historical enterprise financial data through a machine learning algorithm to obtain analysis results, and calculating data quality indicators based on the analysis results;

[0013] Establish data quality metrics based on the calculated data quality indicators.

[0014] Furthermore, the analyzing of historical enterprise financial data by machine learning algorithm includes:

[0015] Perform feature extraction on historical corporate financial data, calculate the importance of the extracted features based on the decision tree algorithm, and select the top k most important features based on the calculation results;

[0016] And perform principal component analysis on the historical corporate financial data characteristics, convert the original feature space into a new feature space through linear transformation, and retain the main features.

[0017] Furthermore, the analyzing of historical enterprise financial data by machine learning algorithm also includes:

[0018] According to the feature importance calculation results and feature principal component analysis results, features are selected as training set data, and the objective function is optimized according to the feature importance calculation results and feature principal component analysis results. The decision tree model is trained based on the optimized objective function and training set data, and the classification labels and scores of the data are output through the decision tree model.

[0019] Furthermore, the data quality indicators of the data quality metric standard include accuracy, completeness, consistency, timeliness and uniqueness; the step S3 specifically includes:

[0020] Step S301, data reception and preprocessing; wherein the data here refers to second data;

[0021] Step S302: Calculate the quality score of each piece of data in real time according to the data quality index;

[0022] Step S303: Calculate the comprehensive quality score of the data using a weighted comprehensive scoring algorithm in combination with multiple data quality indicators;

[0023] Step S304: dynamically adjusting the quality threshold according to the real-time data characteristics through a dynamic threshold adjustment model;

[0024] Step S305: Compare the quality score calculated in real time with the dynamically adjusted threshold, filter out data that meets the conditions, and use it as the third data.

[0025] Furthermore, dynamically adjusting the quality threshold according to the real-time data characteristics through the dynamic threshold adjustment model includes:

[0026] The dynamic threshold adjustment model is trained using historical enterprise financial data, and the dynamic thresholds of various data quality indicators are predicted based on the characteristics of real-time enterprise financial data.

[0027] Furthermore, the step S4 comprises:

[0028] Step S401: Setting access control policies based on different attributes; wherein the different attributes include user roles, user attributes, data attributes, and environment attributes;

[0029] Step S402: Establish a blockchain network, bind each piece of data to its access control policy, and record all changes in access control policies and access control lists on the blockchain;

[0030] Step S403: Use smart contracts to automatically execute access control policies and verify each access request through smart contracts.

[0031] Furthermore, the step S5 comprises:

[0032] Step S501: encrypt the fourth data by homomorphic encryption, and encrypt the homomorphic encryption key by quantum secure encryption to obtain fifth data;

[0033] Step S502: Send the fifth data and the encryption key.

[0034] The second aspect of the present invention discloses an enterprise financial sharing evaluation system based on big data. The system is implemented based on the method disclosed in the first aspect. The system includes a data processing module, an access control module, and a data encryption and transmission module; the data processing module also includes a first data processing submodule, a second data processing submodule, and a third data processing submodule; wherein,

[0035] The first data processing submodule is used to establish a data quality metric standard, and screen the enterprise financial data that needs to be shared based on the data quality metric standard to obtain first data;

[0036] The second data processing submodule is used to perform automatic cleaning and preprocessing operations on the first data to obtain second data;

[0037] The third data processing submodule is used to establish a dynamic data quality evaluation mechanism based on the data quality metric standard, and perform a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data;

[0038] The access control module is used to set access control on the third data to obtain fourth data;

[0039] The data encryption and transmission module is used to perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention analyzes historical corporate financial data based on a decision tree algorithm, establishes data quality metrics that are exclusive to specific data sources and specific data types, and constructs a dynamic data quality assessment mechanism based on the data quality metrics, thereby implementing a two-layer mechanism to screen corporate financial data that needs to be shared, ensuring the data quality and data quality assessment accuracy of corporate financial data sharing assessments, thereby further improving the accuracy of corporate financial data sharing assessments; in addition, the present invention also performs access control and encryption operations on data, further improving data security and privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the economic application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0043] Figure 1 A schematic diagram of the process of the enterprise financial sharing evaluation method based on big data disclosed in an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the structure of an enterprise financial sharing evaluation system based on big data disclosed in yet another embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0046] Embodiment 1

[0047] The first aspect of the present invention discloses a method for evaluating enterprise financial sharing based on big data. Figure 1 , Figure 1: is a flow chart of an enterprise financial sharing evaluation method based on big data disclosed in an embodiment of the present invention, the method comprising the following steps:

[0048] Step S1: establishing a data quality metric standard, and screening the enterprise financial data to be shared based on the data quality metric standard to obtain first data;

[0049] Step S2, performing automated cleaning and preprocessing operations on the first data to obtain second data;

[0050] Step S3: establishing a dynamic data quality evaluation mechanism based on the data quality metric standard, and performing a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data;

[0051] Step S4, setting access control on the third data to obtain fourth data;

[0052] Step S5: perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

[0053] The data cleaning operations in step S2 specifically include deleting duplicate records in the data; using interpolation, mean filling or deleting records with a large number of missing values; and correcting obvious erroneous data according to predefined rules or external reference data. The preprocessing operations specifically include data formatting and standardization.

[0054] In an embodiment of the present invention, by establishing a data quality metric before cleaning and preprocessing the data in step S2, and screening the corporate financial data that needs to be shared based on the data quality metric, so as to screen out high-quality data that needs to be shared, the amount of data that needs to be processed in subsequent steps can be significantly reduced, and the overall processing efficiency can be improved. Based on the data quality metric, it is ensured that the data entering step S2 has met certain quality requirements.

[0055] Furthermore, the operation of establishing a data quality metric standard in step S1 specifically includes analyzing historical enterprise financial data through a machine learning algorithm to obtain analysis results, and calculating data quality indicators based on the analysis results;

[0056] Establish data quality metrics based on the calculated data quality indicators.

[0057] Further, the analysis of historical corporate financial data through machine learning algorithms includes:

[0058] Perform feature extraction on historical corporate financial data, calculate the importance of the extracted features based on the decision tree algorithm, and select the top k most important features based on the calculation results; the calculation process is:

[0059]

[0060] Among them, x j is the jth feature, I j is the jth feature x j The importance of, T is the set of all nodes in the tree, p(t) is the probability of node t, Δi(t) is the information gain of node t, 1 is the indicator function, when the node t segmentation feature is x j When , the value is 1;

[0061] Δi(t)=i(t)-p L i(t L )-p R i(t R )

[0062] i(t) is the information entropy of node t, p R and p L are the probabilities of the left child node and the right child node respectively, i(t L ) and i(t R ) are the information entropy of the left child node and the right child node respectively;

[0063] And perform principal component analysis on the historical corporate financial data characteristics, convert the original feature space into a new feature space through linear transformation, and retain the main features.

[0064] Furthermore, the analysis of historical corporate financial data through machine learning algorithms also includes:

[0065] Select features as training set data according to the feature importance calculation results and feature principal component analysis results, optimize the objective function according to the feature importance calculation results and feature principal component analysis results, train the decision tree model based on the optimized objective function and training set data, and output the classification label and score of the data through the decision tree model;

[0066] Among them, the objective function is:

[0067] O(t)=α·Δi(t)+β·I j +γ·PCA j

[0068] Among them, O(t) is the objective function, PCA j is the jth feature x j The principal component scores in PCA, α, β, and γ are hyperparameters for adjusting weights.

[0069] Specifically, the data quality index is calculated by the classification labels and scores of the data output by the decision tree model. It should be noted that in the historical corporate financial data used for training, the data classification labels and scores have been annotated according to the data quality indicators.

[0070] For example, the classification labels may include A-level quality, B-level quality, and C-level quality. Among them, A-level quality means that the data is complete and has no missing values, high consistency, consistent with records in other data sources, good timeliness, the time interval between data generation and use is within an acceptable range, high uniqueness, and no duplicate records; B-level quality means that the data has a small number of missing values, but does not affect the main analysis, general consistency, a small number of inconsistencies with records in other data sources, general timeliness, a long time interval between data generation and use, but still within an acceptable range, general uniqueness, and a small number of duplicate records; C-level quality means that the data has many missing values, which affects the analysis results, poor consistency, many inconsistencies with records in other data sources, poor data timeliness, the time interval between data generation and use exceeds the acceptable range, poor data uniqueness, and many duplicate records.

[0071] Scoring mainly refers to the scoring of data quality indicators. For example, when we analyze a set of 1,000 records of financial data, we can obtain the number of records with the true label of A-level quality, the number of records with the true label of C-level quality, the number of records predicted by the model as A-level quality and actually as A-level quality, the number of records predicted by the model as C-level quality and actually as C-level quality, the number of records predicted by the model as A-level quality but actually as C-level quality, the number of records predicted by the model as C-level quality but actually as A-level quality, the total number of data records, the number of records with non-missing values, the number of consistent records, the time interval from data generation to use (average number of days), the maximum acceptable time interval (number of days), and the number of non-duplicate records. Based on the above data, we can score and calculate the data quality indicators of this set of data, namely, accuracy score, completeness score, consistency score, timeliness score, and uniqueness score. Among them, the accuracy score can be calculated based on the consistency between the model prediction result and the actual result, the completeness score can be calculated based on the proportion of non-missing values, the consistency score can be calculated based on the proportion of consistent records, the timeliness score can be calculated based on the time interval between data generation and use, and the uniqueness score can be calculated based on the proportion of non-duplicate records.

[0072] Through model analysis, we can obtain the quality of various data quality indicators of historical corporate financial data from different data sources and different data types. Therefore, we can establish data quality metrics that are exclusive to specific data sources and specific data types based on the quality situation. It should be noted that the dynamic data quality assessment mechanism based on data quality metrics is also established based on specific data sources and specific data types.

[0073] Furthermore, the data quality indicators of the data quality metric standard include but are not limited to accuracy, completeness, consistency, timeliness and uniqueness; step S3 specifically includes:

[0074] Step S301, data reception and preprocessing; wherein the data here refers to second data;

[0075] Step S302: Calculate the quality score of each piece of data in real time according to the data quality index:

[0076] Q i (t) = w1·A i (t)+w2·C i (t)+w3·S i (t)+w4·T i (t)+w5·U i (t)

[0077] Among them, Q i (t) is the comprehensive quality score of the i-th data at time t, A i (t), C i (t), S i (t), T i (t), U i (t) are the scores for accuracy, completeness, consistency, timeliness and uniqueness, respectively, and w1~w5 correspond to the weights of each data quality indicator;

[0078] Step S303: Combine multiple data quality indicators and use a weighted comprehensive scoring algorithm to calculate the comprehensive quality score of the data:

[0079]

[0080] Among them, Q(t) is the comprehensive quality score, w i is the comprehensive quality score Q of the i-th data at time t i (t) the corresponding weight;

[0081] Step S304: dynamically adjusting the quality threshold according to the real-time data characteristics through a dynamic threshold adjustment model;

[0082] Step S305: Compare the quality score calculated in real time with the dynamically adjusted threshold value, filter out data that meets the conditions, and use it as the third data, specifically:

[0083] 1(Q i (t))≥θ i (t))

[0084] Among them, 1 is the indicator function, θ i(t) is the dynamic threshold of the i-th data quality indicator at time t. i (t))≥θ i (t) takes the value of 1, otherwise it takes the value of 0.

[0085] Furthermore, dynamically adjusting the quality threshold according to the real-time data characteristics through the dynamic threshold adjustment model includes:

[0086] Use historical enterprise financial data to train a dynamic threshold adjustment model, and predict the dynamic thresholds of various data quality indicators based on the characteristics of real-time enterprise financial data:

[0087] θ i (t) = f ML (X i (t))

[0088] Among them, θ i (t) is the dynamic threshold of the i-th data quality indicator at time t, f ML X is a dynamic threshold adjustment model built by machine learning algorithm. i (t) is the data feature of real-time enterprise financial data.

[0089] Further, step S4 includes:

[0090] Step S401: Setting access control policies based on different attributes; wherein the different attributes include user roles, user attributes, data attributes, and environment attributes;

[0091] Step S402: Establish a blockchain network, bind each piece of data to its access control policy, and record all changes in access control policies and access control lists on the blockchain;

[0092] Step S403: Use smart contracts to automatically execute access control policies and verify each access request through smart contracts.

[0093] In an embodiment of the present invention, by comprehensively considering user roles (such as administrators, ordinary users), user attributes (such as departments, positions), data attributes (such as sensitive data, public data) and environmental attributes (such as access time, geographic location), a more sophisticated access control strategy can be implemented. The access control strategy based on multi-dimensional attributes can more effectively protect data and prevent unauthorized access, and can dynamically adjust the access control strategy according to different scenarios and needs to adapt to changing security requirements and business needs.

[0094] In addition, the transparency and immutability of blockchain are used to manage and record access control policies and access control lists to ensure their security and credibility. The automatic execution characteristics of smart contracts are also used to achieve automated management and verification of access control policies. Each access request is verified in real time through smart contracts to ensure that only requests that meet policy requirements can obtain access rights, thereby improving data processing efficiency and security.

[0095] Further, step S5 includes:

[0096] Step S501: encrypt the fourth data by homomorphic encryption, and encrypt the homomorphic encryption key by quantum secure encryption to obtain fifth data;

[0097] Step S502: Send the fifth data and the encryption key.

[0098] In the embodiments of the present invention, data privacy is ensured through homomorphic encryption operations and calculations are allowed in an encrypted state, which makes it suitable for cloud computing and distributed big data computing environments; secondly, quantum safe encryption operations are used to ensure the security of encryption keys in quantum computing environments, prevent future quantum attacks, and ensure the confidentiality and integrity of data during transmission.

[0099] Embodiment 2

[0100] The second aspect of the present invention discloses an enterprise financial sharing evaluation system based on big data, please refer to Figure 2 , Figure 2 It is a structural diagram of an enterprise financial sharing evaluation system based on big data disclosed in another embodiment of the present invention, the system includes a data processing module, an access control module, and a data encryption and transmission module; the data processing module also includes a first data processing submodule, a second data processing submodule and a third data processing submodule; wherein,

[0101] The first data processing submodule is used to establish a data quality metric standard, and screen the enterprise financial data that needs to be shared based on the data quality metric standard to obtain first data;

[0102] The second data processing submodule is used to perform automatic cleaning and preprocessing operations on the first data to obtain second data;

[0103] The third data processing submodule is used to establish a dynamic data quality evaluation mechanism based on the data quality metric standard, and perform a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data;

[0104] The access control module is used to set access control on the third data to obtain fourth data;

[0105] The data encryption and transmission module is used to perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

[0106] Furthermore, the operation of establishing data quality metrics specifically includes analyzing historical enterprise financial data through machine learning algorithms, obtaining analysis results, and calculating data quality indicators based on the analysis results;

[0107] Establish data quality metrics based on the calculated data quality indicators.

[0108] Further, the analysis of historical corporate financial data through machine learning algorithms includes:

[0109] Perform feature extraction on historical corporate financial data, calculate the importance of the extracted features based on the decision tree algorithm, and select the top k most important features based on the calculation results; the calculation process is:

[0110]

[0111] Among them, x j is the jth feature, I j is the jth feature x j The importance of, T is the set of all nodes in the tree, p(t) is the probability of node t, Δi(t) is the information gain of node t, 1 is the indicator function, when the node t segmentation feature is x j When , the value is 1;

[0112] Δi(t)=i(t)-p L i(t L )-p R i(t R )

[0113] i(t) is the information entropy of node t, p R and p L are the probabilities of the left child node and the right child node respectively, i(t L ) and i(t R ) are the information entropy of the left child node and the right child node respectively;

[0114] And perform principal component analysis on the historical corporate financial data characteristics, convert the original feature space into a new feature space through linear transformation, and retain the main features.

[0115] Furthermore, the analysis of historical corporate financial data through machine learning algorithms also includes:

[0116] Select features as training set data according to the feature importance calculation results and feature principal component analysis results, optimize the objective function according to the feature importance calculation results and feature principal component analysis results, train the decision tree model based on the optimized objective function and training set data, and output the classification label and score of the data through the decision tree model;

[0117] Among them, the objective function is:

[0118] O(t)=α·Δi(t)+β·I j +γ·PCA j

[0119] Among them, O(t) is the objective function, PCA j is the jth feature x j The principal component scores in PCA, α, β, and γ are hyperparameters for adjusting weights.

[0120] Furthermore, the data quality indicators of the data quality metric standard include accuracy, completeness, consistency, timeliness and uniqueness; the third data processing submodule is specifically used to perform the following operations:

[0121] Data reception and preprocessing; wherein the data here refers to the second data;

[0122] Calculate the quality score of each piece of data in real time based on the data quality indicators:

[0123] Q i (t) = w1·A i (t)+w2·C i (t)+w3·S i (t)+w4·T i (t)+w5·U i (t)

[0124] Among them, Q i (t) is the comprehensive quality score of the i-th data at time t, A i (t), C i (t), S i (t), T i (t), U i (t) are the scores for accuracy, completeness, consistency, timeliness and uniqueness, respectively, and w1~w5 correspond to the weights of each data quality indicator;

[0125] Combining multiple data quality indicators, a weighted comprehensive scoring algorithm is used to calculate the comprehensive quality score of the data:

[0126]

[0127] Among them, Q(t) is the comprehensive quality score, wi is the comprehensive quality score Q of the i-th data at time t i (t) the corresponding weight;

[0128] Dynamically adjust the quality threshold according to real-time data characteristics through a dynamic threshold adjustment model;

[0129] The quality score calculated in real time is compared with the dynamically adjusted threshold to screen out the qualified data and use it as the third data, specifically:

[0130] 1(Q i (t))≥θ i (t))

[0131] Among them, 1 is the indicator function, θ i (t) is the dynamic threshold of the i-th data quality indicator at time t. i (t))≥θ i (t) takes the value of 1, otherwise it takes the value of 0.

[0132] Furthermore, dynamically adjusting the quality threshold according to the real-time data characteristics through the dynamic threshold adjustment model includes:

[0133] Use historical enterprise financial data to train a dynamic threshold adjustment model, and predict the dynamic thresholds of various data quality indicators based on the characteristics of real-time enterprise financial data:

[0134] θ i (t) = f ML (X i (t))

[0135] Among them, θ i (t) is the dynamic threshold of the i-th data quality indicator at time t, f ML X is a dynamic threshold adjustment model built by machine learning algorithm. i (t) is the data feature of real-time enterprise financial data.

[0136] Furthermore, the access control module is specifically used to perform the following operations:

[0137] Setting access control policies based on different attributes; wherein the different attributes include user roles, user attributes, data attributes and environment attributes;

[0138] Establish a blockchain network, bind each piece of data to its access control policy, and record all changes in access control policies and access control lists on the blockchain;

[0139] Use smart contracts to automatically enforce access control policies and verify each access request through smart contracts.

[0140] Furthermore, the data encryption and transmission module is specifically used to perform the following operations:

[0141] Encrypting the fourth data by homomorphic encryption, and encrypting the homomorphic encryption key by quantum secure encryption to obtain fifth data;

[0142] The fifth data and the encryption key are sent.

[0143] It should be noted that the specific implementation process of the second embodiment is similar to that of the first embodiment and will not be repeated in the second embodiment.

[0144] Finally, it should be noted that the enterprise financial sharing evaluation method and system based on big data disclosed in the embodiments of the present invention disclose only the preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features therein may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. The enterprise financial sharing evaluation method based on big data is characterized by: The method comprises: Step S1: establishing a data quality metric standard, and screening the enterprise financial data to be shared based on the data quality metric standard to obtain first data; Step S2, performing automated cleaning and preprocessing operations on the first data to obtain second data; Step S3: establishing a dynamic data quality evaluation mechanism based on the data quality metric standard, and performing a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data; Step S4, setting access control on the third data to obtain fourth data; Step S5: perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

2. The enterprise financial sharing evaluation method based on big data according to claim 1 is characterized in that: The operation of establishing a data quality metric standard in step S1 specifically includes analyzing historical enterprise financial data through a machine learning algorithm to obtain analysis results, and calculating data quality indicators based on the analysis results; Establish data quality metrics based on the calculated data quality indicators.

3. The enterprise financial sharing evaluation method based on big data according to claim 2 is characterized in that: The analysis of historical enterprise financial data by machine learning algorithms includes: Perform feature extraction on historical corporate financial data, calculate the importance of the extracted features based on the decision tree algorithm, and select the top k most important features based on the calculation results; And perform principal component analysis on the historical corporate financial data characteristics, convert the original feature space into a new feature space through linear transformation, and retain the main features.

4. The enterprise financial sharing evaluation method based on big data according to claim 3 is characterized in that: The analysis of historical corporate financial data by machine learning algorithms also includes: According to the feature importance calculation results and feature principal component analysis results, features are selected as training set data, and the objective function is optimized according to the feature importance calculation results and feature principal component analysis results. The decision tree model is trained based on the optimized objective function and training set data, and the classification labels and scores of the data are output through the decision tree model.

5. The enterprise financial sharing evaluation method based on big data according to any one of claims 1 to 4, characterized in that: The data quality indicators of the data quality metric standard include accuracy, completeness, consistency, timeliness and uniqueness; the step S3 specifically includes: Step S301, data reception and preprocessing; wherein the data here refers to second data; Step S302: Calculate the quality score of each piece of data in real time according to the data quality index; Step S303: Calculate the comprehensive quality score of the data using a weighted comprehensive scoring algorithm in combination with multiple data quality indicators; Step S304: dynamically adjusting the quality threshold according to the real-time data characteristics through a dynamic threshold adjustment model; Step S305: Compare the quality score calculated in real time with the dynamically adjusted threshold, filter out data that meets the conditions, and use it as the third data.

6. The enterprise financial sharing evaluation method based on big data according to claim 5 is characterized in that: The dynamically adjusting the quality threshold according to the real-time data characteristics by the dynamic threshold adjustment model includes: The dynamic threshold adjustment model is trained using historical enterprise financial data, and the dynamic thresholds of various data quality indicators are predicted based on the characteristics of real-time enterprise financial data.

7. The enterprise financial sharing evaluation method based on big data according to claim 1 is characterized in that: The step S4 comprises: Step S401: Setting access control policies based on different attributes; wherein the different attributes include user roles, user attributes, data attributes, and environment attributes; Step S402: Establish a blockchain network, bind each piece of data to its access control policy, and record all changes in access control policies and access control lists on the blockchain; Step S403: Use smart contracts to automatically execute access control policies and verify each access request through smart contracts.

8. The enterprise financial sharing evaluation method based on big data according to claim 1 is characterized in that: The step S5 comprises: Step S501: encrypt the fourth data by homomorphic encryption, and encrypt the homomorphic encryption key by quantum secure encryption to obtain fifth data; Step S502: Send the fifth data and the encryption key.

9. An enterprise financial sharing evaluation system based on big data, the system is implemented based on the method of any one of claims 1 to 8, characterized in that: The system includes a data processing module, an access control module, and a data encryption and transmission module; the data processing module also includes a first data processing submodule, a second data processing submodule, and a third data processing submodule; wherein, The first data processing submodule is used to establish a data quality metric standard, and screen the enterprise financial data that needs to be shared based on the data quality metric standard to obtain first data; The second data processing submodule is used to perform automatic cleaning and preprocessing operations on the first data to obtain second data; The third data processing submodule is used to establish a dynamic data quality evaluation mechanism based on the data quality metric standard, and perform a data processing operation on the second data based on the dynamic data quality evaluation mechanism to obtain third data; The access control module is used to set access control on the third data to obtain fourth data; The data encryption and transmission module is used to perform an encryption operation on the fourth data to obtain fifth data, and perform data transmission based on the fifth data.

10. The enterprise financial sharing evaluation system based on big data according to claim 9 is characterized in that: The operation of establishing a data quality metric standard specifically includes analyzing historical enterprise financial data through a machine learning algorithm to obtain analysis results, and calculating data quality indicators based on the analysis results; and establishing a data quality metric standard based on the calculated data quality indicators.

Images