Fund exchange information management system based on big data analysis

Abstract

The invention discloses a fund exchange information management system based on big data analysis, and relates to the technical field of fund information management. The fund exchange information management system based on big data analysis comprises a data acquisition module, a preprocessing cleaning module, a data storage management module, a data analysis module, a real-time monitoring anomaly detection module and a risk assessment management module, and the data acquisition module is used for acquiring fund flow data by using a data integration tool; the preprocessing cleaning module is used for preprocessing the capital flow data, the data storage management module is used for storing the capital flow data processed by the preprocessing cleaning module, and the real-time monitoring anomaly detection module is used for monitoring the capital flow data set in real time. And the risk assessment management module is used for performing risk assessment on the fund flow data through the data reliability assessment value. The problem that the reliability and the accuracy of automatic analysis data of the system are insufficient due to data quality problems or algorithm limitations is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of fund information management, and is a fund transaction information management system based on big data analysis. Background Art

[0002] In the modern financial industry, it is essential to manage and monitor the flow of funds efficiently and accurately. Traditional fund flow information management systems rely on a lot of manual operations and simple automation tools, which are often inefficient and error-prone when processing high-speed, large-scale financial data. As the volume of data continues to grow and the speed of transactions increases, the challenges faced by these systems are also increasing.

[0003] On the one hand, the integrity and accuracy of data are often challenged, resulting in reduced reliability of risk assessment and decision support. Existing systems often lack efficient anomaly detection mechanisms and data processing delay optimization, which not only affects operational efficiency but also may bring security risks. In addition, traditional fund flow information management systems may produce inaccurate results in automated data analysis due to data quality issues or algorithm limitations.

[0004] Therefore, there is an urgent need for a fund transaction information management system based on big data analysis to solve the above problems. Summary of the invention

[0005] Technical issues solved

[0006] In view of the deficiencies in the prior art, the present invention provides a fund transaction information management system based on big data analysis, which solves the problem of insufficient reliability and accuracy of system automated analysis data due to data quality issues or algorithm limitations.

[0007] Technical Solution

[0008] The present invention provides the following technical solutions: a fund flow information management system based on big data analysis, comprising a data acquisition module, a preprocessing and cleaning module, a data storage management module, a data analysis module, a real-time monitoring anomaly detection module and a risk assessment management module, wherein the data acquisition module is used to collect fund flow data using a data integration tool; the preprocessing and cleaning module is used to preprocess the fund flow data, record the fund flow data processing delay, and obtain the fund flow data mean absolute error and the fund flow data mean square error; the data storage management module is used to store the fund flow data processed by the preprocessing and cleaning module and obtain the data integrity ratio; the data analysis module is used to predict and classify the fund flow data, obtain a fund flow test data set, and record the correct prediction number and the total prediction number of the fund flow test data set, and obtain the data prediction accuracy by calculating the correct prediction number and the total prediction number of the fund flow test data set; the data accuracy evaluation value is obtained by comprehensive calculation of the fund flow data mean absolute error, the fund flow data mean square error and the data prediction accuracy; the real-time monitoring anomaly detection module is used to use Apache The Kafka distributed stream processing platform monitors the cash flow data set in real time, detects abnormal patterns, obtains the abnormal detection rate, false positive rate and false negative rate of the cash flow data, and obtains the abnormal detection efficiency of the cash flow data through comprehensive calculation of the abnormal detection rate, false positive rate and false negative rate of the cash flow data; the risk assessment management module is used to obtain the data reliability assessment value through comprehensive calculation of the data integrity ratio, the abnormal detection efficiency of the cash flow data and the processing delay of the cash flow data, conducts risk assessment on the cash flow data through the data reliability assessment value, and makes corresponding risk handling and optimization measures.

[0009] Furthermore, the specific process of preprocessing the cash flow data, recording the processing delay of the cash flow data, and obtaining the average absolute error of the cash flow data and the mean square error of the cash flow data is as follows: cleaning, formatting and standardizing the cash flow data collected from the data acquisition module, recording the actual observation value of the data, the processed data value and the total number of points of the preprocessed data; obtaining the average absolute error of the cash flow data by comprehensively calculating the actual observation value of the data, the processed data value and the total number of points of the preprocessed data; obtaining the mean square error of the cash flow data by comprehensively calculating the actual observation value of the data, the processed data value and the total number of points of the preprocessed data.

[0010] Furthermore, the specific calculation formula for the mean absolute error of capital flow data is: In the formula, MAE represents the mean absolute error of fund flow data, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data value.

[0011] Furthermore, the specific calculation formula of the mean square error of the capital flow data is: In the formula, MSE represents the mean square error of fund flow data, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data value.

[0012] Furthermore, the specific process of storing the cash flow data processed by the pre-processing cleaning module and obtaining the data integrity ratio is as follows: using Apache Cassandra and Amazon S3 to store the cash flow data, recording the number of damaged and missing data points and the total number of data points stored in the database during the storage process; and obtaining the data integrity ratio by comprehensively calculating the number of damaged and missing data points and the total number of data points stored in the database.

[0013] Furthermore, the specific calculation formula of the data integrity ratio is: In the formula, DI represents the data integrity ratio, NCD represents the number of damaged and missing data points, and TDP represents the total number of data points stored in the database.

[0014] Furthermore, the specific calculation formula of the data accuracy assessment value is: AAV=w1(1-MAE)+w2(1-MSE)+w3ACC; wherein AAV represents the data accuracy assessment value, MAE represents the mean absolute error of cash flow data, w1 represents the weight factor of the mean absolute error of cash flow data, MSE represents the mean square error of cash flow data, w2 represents the weight factor of the mean square error of cash flow data, ACC represents the data prediction accuracy, and w3 represents the weight factor of the data prediction accuracy.

[0015] Furthermore, the specific calculation formula for the abnormal detection efficiency of capital flow data is: Where ADE represents the anomaly detection efficiency of capital flow data, ADR represents the anomaly detection rate of capital flow data, FPR represents the false positive rate, FNR represents the false negative rate, and ∈ is a constant.

[0016] Furthermore, the specific calculation formula of the data reliability assessment value is: RAV=β1·DI+β2·ADE+β3·DPL; wherein RAV represents the data reliability assessment value, DI represents the data integrity ratio, β1 represents the weight coefficient of the data integrity ratio, ADE represents the anomaly detection efficiency of cash flow data, β2 represents the weight coefficient of the anomaly detection efficiency of cash flow data, DPL represents the processing delay of cash flow data, and β3 represents the weight coefficient of the processing delay of cash flow data.

[0017] Furthermore, the risk assessment of the capital flow data is performed through the data reliability assessment value, and corresponding risk handling and optimization measures are made. The specific process is: compare the data reliability assessment value with the threshold in real time, and when the data reliability assessment value is greater than the threshold, no processing is performed; when the data reliability assessment value is less than or equal to the threshold, the capital flow data collected in the data acquisition module is repeatedly cleaned, formatted and standardized, and the processing intensity of Apache Kafka distributed stream processing in the real-time monitoring anomaly detection module is strengthened until the reliability assessment value is increased above the threshold.

[0018] Beneficial Effects

[0019] The present invention has the following beneficial effects:

[0020] (1) The fund flow information management system based on big data analysis pre-processes the fund flow data and records the data processing delay through the pre-processing and cleaning module, and obtains the mean absolute error and mean square error of the fund flow data, thereby achieving a significant improvement in data processing efficiency and accuracy, and effectively solving the problem of slow and error-prone data processing in the prior art.

[0021] (2) The fund flow information management system based on big data analysis uses a data storage management module to store pre-processed and cleaned fund flow data and obtain a data integrity ratio, thereby achieving a high degree of data integrity and reliability, and effectively solving the problems in the prior art that data is easily damaged and lost, and it is difficult to ensure data quality.

[0022] (3) The fund flow information management system based on big data analysis uses the Apache Kafka distributed stream processing platform to monitor the fund flow data set in real time through a real-time monitoring anomaly detection module, and detects abnormal patterns. The anomaly detection efficiency of the fund flow data is obtained by comprehensively calculating the anomaly detection rate, false positive rate, and false negative rate of the fund flow data, thereby optimizing risk management, ensuring the high efficiency of anomaly detection, and effectively solving the problems of slow abnormal response and high false alarm rate in the existing technology.

[0023] (4) The fund flow information management system based on big data analysis predicts and classifies fund flow data through the data analysis module and calculates the accuracy of the fund flow test data set, thereby achieving the depth and breadth of data analysis, effectively improving the decision-making support capabilities of the system, and solving the problem of lack of effective decision-making and analysis tools in the existing technology.

[0024] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1This is a structural diagram of a fund transaction information management system based on big data analysis in the present invention. DETAILED DESCRIPTION

[0026] In conjunction with the drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described. It should be noted that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0027] See also Figure 1 The embodiment of the present invention provides a technical solution: a fund flow information management system based on big data analysis, including a data acquisition module, a preprocessing and cleaning module, a data storage management module, a data analysis module, a real-time monitoring anomaly detection module and a risk assessment management module, wherein the data acquisition module is used to collect fund flow data using a data integration tool; the preprocessing and cleaning module is used to preprocess the fund flow data, record the fund flow data processing delay, and obtain the fund flow data mean absolute error and the fund flow data mean square error; the data storage management module is used to store the fund flow data processed by the preprocessing and cleaning module and obtain the data integrity ratio; the data analysis module is used to predict and classify the fund flow data, obtain a fund flow test data set, and record the correct prediction number and the total prediction number of the fund flow test data set, and calculate the fund flow test data set. The data prediction accuracy is calculated by the ratio of the number of correct predictions to the total number of predictions; the data accuracy assessment value is obtained by comprehensive calculation of the mean absolute error of the capital flow data, the mean square error of the capital flow data and the data prediction accuracy; the real-time monitoring anomaly detection module is used to use the Apache Kafka distributed stream processing platform to monitor the capital flow data set in real time, detect abnormal patterns, obtain the anomaly detection rate, false positive rate and false negative rate of the capital flow data, and obtain the anomaly detection efficiency of the capital flow data by comprehensive calculation of the anomaly detection rate, false positive rate and false negative rate of the capital flow data; the risk assessment management module is used to obtain the data reliability assessment value by comprehensive calculation of the data integrity ratio, the anomaly detection efficiency of the capital flow data and the processing delay of the capital flow data, conduct risk assessment on the capital flow data through the data reliability assessment value, and make corresponding risk handling and optimization measures.

[0028] Specifically, the cash flow data is preprocessed, the processing delay of the cash flow data is recorded, and the mean absolute error and mean square error of the cash flow data are obtained. The specific process is as follows: In the data acquisition module, the cash flow data first undergoes a series of fine cleaning, formatting and standardization processes to ensure the consistency and accuracy of subsequent analysis. This process includes the screening and error correction of data collected from various data sources (such as bank transaction systems, online payment platforms, etc.), removing invalid and duplicate data entries, and converting all data into a unified format. At the same time, the system will record in detail the actual observation value of each data, the processed data value, and the total number of data points involved in the processing process. These records are crucial for subsequent data quality control and auditing.

[0029] After the data preprocessing is completed, the system uses the recorded actual observations and processed data values ​​to calculate the mean absolute error of the capital flow data through a certain mathematical formula. The mean absolute error is an indicator that measures the difference between the predicted value and the actual observed value. It is extremely important for evaluating the accuracy of the data processing process. Similarly, the system will also calculate the mean square error of the capital flow data.

[0030] In this implementation scheme, in the data preprocessing and cleaning module, the capital flow data is processed through a series of refined processing processes, including cleaning, formatting and standardization, which significantly improves the quality and consistency of the data. These steps ensure that the data collected from various data sources undergoes strict error checking and redundancy elimination, minimizing errors and inconsistencies in data processing. By recording the actual observed values ​​of the data and the processed data values ​​in detail, and calculating the mean absolute error and mean square error, the system can effectively evaluate the accuracy and reliability of data processing. This process not only enhances the accuracy of data analysis, but also provides solid data support for risk assessment and decision-making. Therefore, the implementation of this step greatly improves the operational efficiency and data credibility of the entire capital flow information management system, and provides strong data support for financial institutions, enabling them to more effectively manage capital flows, predict and mitigate potential risks. Such an effect effectively solves the defects of traditional systems in data consistency and accuracy, thereby enhancing the overall performance of the entire system.

[0031] Specifically, the calculation formula for the mean absolute error of fund flow data is: In the formula, MAE represents the mean absolute error of fund flow data, which is a key indicator to measure the accuracy of data processing, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data value. This statistical method evaluates the effect of the data preprocessing step by calculating the average of the absolute differences between the actual observations and the processed data values.

[0032] In this implementation scheme, the mean absolute error of fund flow data provides an intuitive and easy-to-understand data quality assessment, helping to identify and improve possible problems in the data processing process, thereby improving the data processing and analysis capabilities of the entire fund management system.

[0033] Specifically, the calculation formula for the mean square error of capital flow data is: In the formula, MSE represents the mean square error of capital flow data, which is an important indicator to measure the accuracy of data preprocessing, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data values. By calculating the average of the sum of squares of the differences between the actual observations and the processed data values, it ensures that all processed data points are included in the calculation, thus providing a comprehensive and consistent error assessment.

[0034] In this implementation scheme, the calculation of the mean square error of cash flow data provides a key quantitative indicator for the data processing accuracy of the system. By calculating the mean square error of cash flow data, the system can evaluate in detail the difference between the actual observed value and the processed data value, thereby identifying and quantifying the size of the error in the data preprocessing step. Specifically, the mean square error of cash flow data emphasizes the impact of large errors in the form of squared differences, allowing the system to more sensitively capture and process abnormal data points that may have a significant impact on the final data analysis results. This calculation step not only helps improve data processing algorithms and ensure data quality, but also plays a decisive role in optimizing the performance and reliability of the entire system. By regularly and systematically evaluating the mean square error of cash flow data, the management team can continuously monitor the effectiveness of the data processing process and adjust the processing strategy accordingly, ultimately achieving the dual effects of improving data analysis accuracy and reducing risk judgment errors.

[0035] Specifically, the cash flow data processed by the pre-processing and cleaning module is stored to obtain the data integrity ratio. The specific process is as follows: Apache Cassandra and Amazon S3 are used to store the cash flow data. Both solutions provide highly reliable and scalable data storage capabilities, suitable for processing large amounts of data; Apache Cassandra provides high-performance distributed database support, while Amazon S3 is known for its durability and object storage capabilities, ensuring long-term and secure data storage. During the storage process, the system will record in detail whether each data point is complete, including the number of damaged or missing data points. At the same time, the system will also record the total number of data points stored in the database, including all successfully stored data points. Through these records, the system can comprehensively calculate the data integrity ratio.

[0036] In this implementation, Apache Cassandra and Amazon S3 are used to store fund flow data, which significantly improves the storage efficiency and reliability of data. By combining these two technologies, the system not only ensures high availability and scalability of data storage, but also can handle large-scale data streams from different data sources. During the storage process, the system accurately records the number of damaged or missing data points and the total number of successfully stored data points. This information is used to calculate the data integrity ratio, which is an indicator of data integrity and accuracy. This ratio is calculated by calculating the ratio of damaged or missing data points to total data points, which directly reflects the quality and reliability of the storage system.

[0037] Specifically, the data integrity ratio is calculated as follows: In the formula, DI represents the data integrity ratio, which is used to measure the degree of data integrity during data storage. NCD represents the number of damaged and missing data points, including all data points that were not correctly stored in the database due to various reasons (such as transmission errors, storage failures, or data corruption). TDP represents the total number of data points stored in the database, that is, all data points that were attempted to be stored. Through this ratio, the system can quantify the degree of data loss during the storage process, thereby evaluating the performance and reliability of the data storage system. The higher the calculated data integrity ratio, the better the data integrity and accuracy, and the higher the reliability of the system.

[0038] In this implementation, by calculating the data integrity ratio, the integrity and accuracy of the data during the storage process can be effectively monitored and evaluated. The level of the data integrity ratio directly reflects the reliability and efficiency of the data storage system. A high data integrity ratio means less data damage and loss, showing the efficiency and stability of the system. By tracking this indicator in real time, the system can immediately identify data damage and missing problems, so as to quickly take repair measures to ensure data quality and integrity.

[0039] Specifically, the data accuracy assessment value is calculated as follows: AAV = w1(1-MAE) + w2(1-MSE) + w3ACC; where AAV represents the data accuracy assessment value, which is used to comprehensively measure the accuracy in the data processing and analysis process, MAE represents the mean absolute error of the capital flow data, w1 represents the weight factor of the mean absolute error of the capital flow data, MSE represents the mean square error of the capital flow data, w2 represents the weight factor of the mean square error of the capital flow data, ACC represents the data prediction accuracy, and w3 represents the weight factor of the data prediction accuracy. The value range of the weight factors w1, w2 and w3 is usually between 0 and 1, and these weights reflect the relative importance of each indicator in the total evaluation value. The specific values ​​of these weight factors can be determined through historical data analysis, expert opinions or through optimization algorithms to ensure the accuracy and practicality of the evaluation model. If the prediction accuracy is considered to be the most important for system stability, then w3 may be set higher than w1 and w2.

[0040] In this implementation scheme, the calculation of the data accuracy assessment value is achieved through a carefully designed formula that combines the mean absolute error, mean square error and data prediction accuracy of the cash flow data. These three indicators are weighted by the corresponding weight factors w1, w2 and w3, reflecting their importance in the overall data accuracy assessment. The setting of these weight factors is based on historical data analysis, business needs assessment and optimization algorithms to ensure the practicality and accuracy of the evaluation results. In this way, the data accuracy assessment value provides a quantitative and comprehensive perspective to evaluate the performance of the system in data processing and analysis. This not only helps to identify and optimize deficiencies in the data processing process, but also enhances the confidence and basis for data-driven decision-making. Therefore, this step effectively improves the data processing quality and decision-making support capabilities of the entire system, making cash flow management more efficient and accurate.

[0041] Specifically, the calculation formula for the abnormal detection efficiency of capital flow data is: Where ADE represents the anomaly detection efficiency of capital flow data, which is a key indicator to measure the performance of anomaly detection system. ADR represents the anomaly detection rate of capital flow data, which reflects the proportion of anomalies successfully identified by the system. FPR represents the false positive rate, FNR represents the false negative rate, and ∈ is a constant, which is used to avoid the situation where the denominator is zero and ensure the stability of calculation.

[0042] In this implementation, through this formula, the system can quantify the effect of anomaly detection, reflecting the performance of the detection model in actual operation. High anomaly detection efficiency of fund flow data indicates that the anomaly detection system can not only efficiently identify abnormal transactions, but also perform well in reducing false positives and false negatives. This measurement is crucial for optimizing risk management strategies because it directly affects the speed and accuracy of response to potential risks, thus providing great support in real-time monitoring and security protection.

[0043] Specifically, the data reliability assessment value is calculated as follows: RAV = β1·DI + β2·ADE + β3·DPL; where RAV represents the data reliability assessment value, which comprehensively considers the integrity of the data, the efficiency of anomaly detection and the processing delay, and provides a quantitative assessment of the overall data reliability of the system. DI represents the data integrity ratio, which reflects the integrity and accuracy of the data during storage. β1 represents the weight coefficient of the data integrity ratio. ADE represents the efficiency of anomaly detection of cash flow data, which measures the ability of the system to identify and respond to abnormal events. β2 represents the weight coefficient of the efficiency of anomaly detection of cash flow data. DPL represents the delay in processing cash flow data, which indicates the speed and timeliness of data processing. β3 represents the weight coefficient of the delay in processing cash flow data. The value range of these weight factors is usually 0 to 1. Their specific values ​​can be set according to the specific needs and priorities of the system, and are usually obtained through historical data analysis, expert experience or optimization algorithms. For example, if the system has extremely high requirements for data integrity, then β1 may be set to a higher value. If the system has more critical requirements for real-time performance, β3 may be correspondingly higher.

[0044] In this embodiment, by calculating the data reliability evaluation value, the system provides managers with a comprehensive metric by which they can evaluate and monitor the performance of the system in key aspects, including data integrity, anomaly detection efficiency, and cash flow data processing latency. The data reliability evaluation value formula effectively quantifies the overall data processing quality and efficiency of the system by combining these three indicators with their respective weight factors (β1, β2, β3). The setting of these weight factors is based on a detailed analysis of system operation and a deep understanding of business needs, ensuring that the importance of each aspect is appropriately reflected. Therefore, the calculation of the data reliability evaluation value not only helps system administrators identify and prioritize key issues in the data management process, but also promotes continuous improvement and optimization of the system. This comprehensive data reliability assessment ensures the efficiency and reliability of the system when processing complex and large-scale cash flow data, providing strong support for financial institutions to cope with rapidly changing market conditions and high standards of compliance requirements.

[0045] Specifically, the risk assessment of capital flow data is carried out through the data reliability assessment value, and corresponding risk treatment and optimization measures are made. The specific process is as follows: First, the system compares the data reliability assessment value with the preset threshold in real time. This threshold is precisely set based on historical data analysis and business needs to define the minimum acceptance standard for data reliability. When the data reliability assessment value is greater than this threshold, it indicates that the performance of the data processing and storage process meets or exceeds the expected standard, and the system will not perform additional processing. On the contrary, when the data reliability assessment value is less than or equal to the threshold, it indicates that there is a potential risk in the system's data processing. At this time, the system will automatically initiate a series of risk treatment and optimization measures. The specific operation includes re-executing the cleaning, formatting and standardization of the capital flow data in the data acquisition module. In addition, in order to improve the efficiency and accuracy of data monitoring, the system will strengthen the processing intensity of Apache Kafka distributed stream processing in the real-time monitoring anomaly detection module to ensure that all data flows are efficiently and strictly checked. These measures will continue to be implemented until the data reliability assessment value is raised above the threshold again. This dynamic adjustment process ensures that the system can continuously monitor and adjust its performance to cope with the ever-changing data quality and system requirements, thereby maintaining the efficient operation and data security of the entire capital flow information management system.

[0046] In this implementation scheme, by real-time monitoring and adjusting the data reliability assessment value relative to the preset threshold, the system can effectively perform risk assessment and management to ensure that the quality of data processing and storage always meets or exceeds business requirements. When the data reliability assessment value is higher than the threshold, the system maintains the current operating state, which means that the data integrity, anomaly detection efficiency and processing speed are all within an acceptable range without additional intervention. Once the data reliability assessment value drops below the threshold, the system immediately initiates a series of optimization and correction measures, including re-execution of data cleaning, formatting and standardization, while enhancing the intensity of real-time data monitoring. This not only quickly improves the processing quality of data, but also enhances the system's ability to respond to potential risks. This dynamic risk management strategy enables the system to minimize data processing errors and risk exposure while maintaining efficient operation, thereby ensuring the stability and reliability of the entire cash flow management system.

[0047] 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", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only the mentioned elements, but also other elements not explicitly listed, or includes elements inherent to such process, method, article or device.

[0048] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all details in detail, nor are they limited to the only specific implementation methods of the invention. Obviously, according to the content of this specification, there is still room for subsequent modifications and changes. The embodiments selected and specifically described in this specification are to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A fund transaction information management system based on big data analysis, comprising a data acquisition module, a pre-processing and cleaning module, a data storage management module, a data analysis module, a real-time monitoring anomaly detection module and a risk assessment management module, characterized in that: The data collection module is used to collect fund flow data using a data integration tool; The preprocessing and cleaning module is used to preprocess the capital flow data, record the capital flow data processing delay, and obtain the mean absolute error and mean square error of the capital flow data; The data storage management module is used to store the capital flow data processed by the pre-processing cleaning module and obtain the data integrity ratio; The data analysis module is used to predict and classify the capital flow data, obtain a capital flow test data set, and record the correct prediction number and the total prediction number of the capital flow test data set, and obtain the data prediction accuracy by calculating the proportion of the correct prediction number and the total prediction number of the capital flow test data set; The data accuracy assessment value is obtained by comprehensively calculating the mean absolute error of the capital flow data, the mean square error of the capital flow data and the data prediction accuracy; The real-time monitoring anomaly detection module is used to use the Apache Kafka distributed stream processing platform to monitor the capital flow data set in real time, detect abnormal patterns, obtain the capital flow data anomaly detection rate, false positive rate and false negative rate, and obtain the capital flow data anomaly detection efficiency by comprehensively calculating the capital flow data anomaly detection rate, false positive rate and false negative rate; The risk assessment management module is used to obtain a data reliability assessment value through comprehensive calculation of the data integrity ratio, cash flow data anomaly detection efficiency and cash flow data processing delay, conduct risk assessment on the cash flow data through the data reliability assessment value, and make corresponding risk handling and optimization measures.

2. A fund transaction information management system based on big data analysis according to claim 1, characterized in that: The specific process of preprocessing the capital flow data, recording the capital flow data processing delay, and obtaining the mean absolute error and mean square error of the capital flow data is as follows: Clean, format and standardize the fund flow data collected from the data collection module, and record the actual observation value of the data, the value of the processed data and the total number of pre-processed data points; The mean absolute error of fund flow data is obtained by comprehensive calculation of the actual observed data value, the processed data value and the total number of pre-processed data points; The mean square error of the capital flow data is obtained by comprehensive calculation of the actual observation value of the data, the processed data value and the total number of points of the pre-processed data.

3. A fund transaction information management system based on big data analysis according to claim 2, characterized in that: The specific calculation formula for the mean absolute error of the capital flow data is: In the formula, MAE represents the mean absolute error of fund flow data, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data value.

4. A fund transaction information management system based on big data analysis according to claim 2, characterized in that: The specific calculation formula for the mean square error of the capital flow data is: In the formula, MSE represents the mean square error of fund flow data, n represents the total number of preprocessed data points, and y i represents the actual observed value of the data, Represents the processed data value.

5. The fund transaction information management system based on big data analysis according to claim 1 is characterized by: The specific process of storing the capital flow data processed by the pre-processing cleaning module and obtaining the data integrity ratio is as follows: Use Apache Cassandra and Amazon S3 to store fund flow data, record the number of data points that are damaged or missing during the storage process and the total number of data points stored in the database; The data integrity ratio is calculated by combining the number of damaged and missing data points with the total number of data points stored in the database.

6. A fund transaction information management system based on big data analysis according to claim 5, characterized in that: The specific calculation formula of the data integrity ratio is: In the formula, DI represents the data integrity ratio, NCD represents the number of damaged and missing data points, and TDP represents the total number of data points stored in the database.

7. The fund transaction information management system based on big data analysis according to claim 1 is characterized by: The specific calculation formula of the data accuracy evaluation value is: AAV=w1(1-MAE)+w2(1-MSE)+w3ACC; In the formula, AAV represents the data accuracy assessment value, MAE represents the mean absolute error of fund flow data, w1 represents the weight factor of the mean absolute error of fund flow data, MSE represents the mean square error of fund flow data, w2 represents the weight factor of the mean square error of fund flow data, ACC represents the data prediction accuracy, and w3 represents the weight factor of the data prediction accuracy.

8. The fund transaction information management system based on big data analysis according to claim 1, characterized in that: The specific calculation formula for the abnormal detection efficiency of capital flow data is: Where ADE represents the anomaly detection efficiency of capital flow data, ADR represents the anomaly detection rate of capital flow data, FPR represents the false positive rate, FNR represents the false negative rate, and ∈ is a constant.

9. The fund transaction information management system based on big data analysis according to claim 1, characterized in that: The specific calculation formula of the data reliability evaluation value is: RAV=β1·DI+β2·ADE+β3·DPL; In the formula, RAV represents the data reliability assessment value, DI represents the data integrity ratio, β1 represents the weight coefficient of the data integrity ratio, ADE represents the anomaly detection efficiency of capital flow data, β2 represents the weight coefficient of the anomaly detection efficiency of capital flow data, DPL represents the processing delay of capital flow data, and β3 represents the weight coefficient of the processing delay of capital flow data.

10. The fund transaction information management system based on big data analysis according to claim 1, characterized in that: The specific process of conducting risk assessment on fund flow data through data reliability assessment value and making corresponding risk treatment and optimization measures is as follows: Compare the data reliability evaluation value with the threshold in real time. If the data reliability evaluation value is greater than the threshold, no processing will be performed. When the data reliability assessment value is less than or equal to the threshold, the capital flow data collected in the data collection module is repeatedly cleaned, formatted and standardized, and the processing intensity of Apache Kafka distributed stream processing in the real-time monitoring anomaly detection module is strengthened until the reliability assessment value is increased above the threshold.

Images