A financial information processing method and system based on artificial intelligence

Abstract

The present invention proposes a financial information processing method and system based on artificial intelligence. It belongs to the field of artificial intelligence technology. The method includes: obtaining financial information data from a financial system, encrypting the financial information data, compressing the encrypted financial information data and transmitting it to a blockchain network; the blockchain node processes the financial information data accordingly after receiving it, and stores the processed financial information data on an alliance chain node. The alliance chain node calculates and processes the data; after obtaining the data from the financial system, the data is encrypted and compressed to ensure the security and integrity of the data during transmission.

Description

Technical Field

[0001] The present invention proposes a financial information processing method and system based on artificial intelligence, belonging to the technical field of artificial intelligence. Background Art

[0002] With the development of science and technology, artificial intelligence technology has been widely used in various fields. In the financial field, the traditional financial information processing method has problems such as low data security and low processing efficiency. Therefore, how to use artificial intelligence technology to process financial information efficiently and safely has become a hot topic in current research.

[0003] At present, some financial information processing methods based on artificial intelligence have appeared on the market, but these methods still have some problems in data transmission, storage and processing. For example, there is a risk of data leakage during data transmission, low data storage security, low data processing efficiency and other issues. In addition, these methods often lack in-depth mining and utilization of data during the processing process, resulting in inaccurate and incomplete processing results. Summary of the invention

[0004] The present invention provides a financial information processing method and system based on artificial intelligence to solve the problems mentioned in the above background technology:

[0005] The present invention proposes a financial information processing method based on artificial intelligence, the method comprising:

[0006] Obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network;

[0007] The blockchain node processes the financial information data accordingly after receiving it, and stores the processed financial information data on the alliance chain node, and performs calculations on the data through the alliance chain node;

[0008] Collect the processing result data after calculation and processing by each node in the alliance chain, integrate the processing result data, extract features from the integrated processing result data, and form a data set; input the data set into the artificial intelligence algorithm model to obtain a prediction model;

[0009] Deploy the prediction model into the financial system and predict the company's financial status by obtaining financial information data in real time.

[0010] Furthermore, the method of obtaining financial information data from a financial system, encrypting the financial information data, and compressing the encrypted financial information data and transmitting it to a blockchain network includes:

[0011] Export the original financial information data that needs to be transmitted from the financial system and back up the exported financial information data;

[0012] Encrypt the financial information data using the AES encryption algorithm, and compress the encrypted financial information data using the compression algorithm;

[0013] Balancing the compression rate and compression speed of the compressed financial information data, and after the compression is completed, performing an integrity check on the data;

[0014] Split the integrity-checked data into multiple data slices, determine the optimal slice size and number based on network bandwidth and latency, and assign a unique identifier to each slice;

[0015] A multi-channel transmission protocol is used to transmit data slices to the blockchain network through a secure network channel, and the transmission parameters are dynamically adjusted according to the network conditions.

[0016] Furthermore, the blockchain node processes the financial information data accordingly after receiving it, and stores the processed financial information data on the alliance chain node, and the alliance chain node performs calculation processing on the data; including:

[0017] After receiving the financial information data, the blockchain node decrypts the financial information data using a decryption algorithm and decompresses the decrypted financial information data;

[0018] The decrypted and decompressed financial information data is stored in the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of the financial information data;

[0019] Dividing the financial information data stored in different sub-nodes into multiple data blocks, and processing the data blocks by a parallel processing algorithm;

[0020] Monitor the computing resource load of each sub-node in real time during the processing process, use load balancing algorithm and resource scheduler to adjust the computing resources between each sub-node in real time;

[0021] After the processing is completed, the processing result data of each child node is obtained.

[0022] Furthermore, the processing result data after calculation and processing by each node in the alliance chain is collected, and the processing result data is integrated, and the feature extraction is performed on the integrated processing result data to form a data set; the data set is input into the artificial intelligence algorithm model to obtain a prediction model; including:

[0023] Collect the processing result data after calculation from each node in the alliance chain, and send the result data to the federated learning server, and integrate the result data through the federated learning server;

[0024] Extracting features from the integrated data and selecting feature variables, combining the extracted features to form a data set, preprocessing the data set, and dividing the preprocessed data set into a training set, a test set, and a validation set;

[0025] Select an AI algorithm model based on the problem type and data characteristics, initialize the model parameters, send the data set to each node through the federated learning server, train the model with the training set, test the trained model with the test set, and verify the performance and generalization ability of the model with the validation set;

[0026] After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters.

[0027] After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0028] Furthermore, the prediction model is deployed into the financial system, and the company's financial status is predicted by acquiring financial information data in real time; including:

[0029] Obtain the required financial information data from the financial system in real time, pre-process the financial information data, and predict the pre-processed real-time data through the prediction model to obtain the prediction results of the company's financial status;

[0030] The real-time prediction results are displayed in a visual way, and the real-time performance of the prediction model is monitored and optimized and adjusted as needed.

[0031] The present invention proposes a financial information processing system based on artificial intelligence, the system comprising:

[0032] Data transmission module: obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network;

[0033] Data processing module: after receiving the financial information data, the blockchain node processes it accordingly, stores the processed financial information data on the alliance chain node, and performs calculations on the data through the alliance chain node;

[0034] Model acquisition module: The model collects the processing result data after calculation and processing by each node in the alliance chain, integrates the processing result data, extracts features from the integrated processing result data, and forms a data set; the data set is input into the artificial intelligence algorithm model to obtain a prediction model;

[0035] Real-time prediction module: deploy the prediction model into the financial system and predict the company's financial status by acquiring financial information data in real time.

[0036] Furthermore, the data transmission module includes:

[0037] Information acquisition module: export the original financial information data to be transmitted from the financial system and back up the exported financial information data;

[0038] Data compression module: encrypts financial information data using the AES encryption algorithm, and compresses the encrypted financial information data using the compression algorithm;

[0039] Data checking module: balances the compression rate and compression speed of the compressed financial information data, and performs integrity check on the data after compression;

[0040] Identifier allocation module: splits the integrity-checked data into multiple data slices, determines the optimal slice size and number based on network bandwidth and latency, and assigns a unique identifier to each slice;

[0041] Multi-channel transmission module: uses a multi-channel transmission protocol to transmit data slices to the blockchain network through a secure network channel, and dynamically adjusts transmission parameters according to network conditions;

[0042] Furthermore, the data processing module includes:

[0043] Data decryption module: after receiving the financial information data, the blockchain node decrypts the financial information data through a decryption algorithm and decompresses the decrypted financial information data;

[0044] Multi-node storage module: storing the decrypted and decompressed financial information data on the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of the financial information data;

[0045] Parallel processing module: divides the financial information data stored in different sub-nodes into multiple data blocks, and processes the data blocks through parallel processing algorithms;

[0046] Resource balancing module: monitors the computing resource load of each sub-node in real time, uses load balancing algorithm and adjusts the computing resources between each sub-node in real time through resource scheduler;

[0047] Data acquisition module: After processing, obtain the processing result data of each sub-node.

[0048] Furthermore, the model acquisition module includes:

[0049] Data integration module: collects the processing result data after calculation from each node in the alliance chain, and sends the result data to the federated learning server, and integrates the result data through the federated learning server;

[0050] Feature extraction module: extract features from the integrated data, select feature variables, combine the extracted features to form a data set, preprocess the data set, and divide the preprocessed data set into a training set, a test set, and a validation set;

[0051] Model selection module: selects an AI algorithm model based on the problem type and data characteristics, initializes the model parameters, sends the data set to each node through the federated learning server, trains the model with the training set, tests the trained model with the test set, and verifies the performance and generalization ability of the model with the validation set;

[0052] Parameter aggregation module: After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters.

[0053] Model optimization module: After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0054] Furthermore, the real-time prediction module includes:

[0055] Status acquisition module: obtains the required financial information data from the financial system in real time, pre-processes the financial information data, predicts the pre-processed real-time data through the prediction model, and obtains the prediction result of the company's financial status;

[0056] Visualization display module: Displays real-time prediction results in a visual way, monitors the real-time performance of the prediction model, and optimizes and adjusts it as needed.

[0057] The beneficial effects of the present invention are as follows: after obtaining data from the financial system, the data is encrypted and compressed to ensure the security and integrity of the data during transmission. The AES encryption algorithm and multi-channel transmission protocol are used to further enhance the security of the data; on the blockchain node, after the data is decrypted and decompressed, it is processed by a parallel processing algorithm, while the computing resource load is monitored in real time, and the load balancing algorithm is used to adjust the resources, thereby improving the speed and efficiency of data processing; by adopting the federated learning technology, the processed data is collected from each node in the alliance chain for feature extraction and model training. This method realizes the training and optimization of the model without leaking the original data, and protects the privacy of each node; the prediction model deployed to the financial system can obtain financial information data in real time, predict the financial status, and display the prediction results in a visual way. This helps the company to understand the financial status in a timely manner and make decisions; the method is based on blockchain and federated learning technology, so that the system has good flexibility and scalability, and nodes can be easily added or deleted to meet the development needs of the company; the financial information data is processed and predicted by the artificial intelligence algorithm model, which realizes the automation and intelligence of financial processing and improves work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a step diagram of the method of the present invention;

[0059] Figure 2 This is a system module diagram of the present invention. DETAILED DESCRIPTION

[0060] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0061] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. The embodiments described are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0063] One embodiment of the present invention, as Figure 1 As shown, a financial information processing method based on artificial intelligence, the method comprising:

[0064] Obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network;

[0065] After receiving the financial information data, the blockchain node processes the data accordingly, and stores the processed financial information data on the alliance chain node, and performs computation on the data through the alliance chain node; the corresponding processing includes decryption and decompression;

[0066] Collect the processing result data after calculation and processing by each node in the alliance chain, integrate the processing result data, extract features from the integrated processing result data, and form a data set; input the data set into the artificial intelligence algorithm model to obtain a prediction model;

[0067] Deploy the prediction model into the financial system and predict the company's financial status by obtaining financial information data in real time.

[0068] The working principle of the above technical solution is as follows: first, the system obtains financial information data from the financial system, and then encrypts the data to ensure data security and privacy protection; the encrypted financial information data is compressed to reduce the bandwidth and storage space required for data transmission. Then, the compressed data is transmitted to the blockchain network to ensure the traceability and tamper-proofness of the data; the blockchain node that receives the financial information data decrypts and decompresses it accordingly to restore the original data. The processed financial information data is stored on the alliance chain node, which has high credibility and security; each node in the alliance chain performs calculations on the processed financial information data, such as statistics and analysis. The processing result data is collected and integrated to form a comprehensive data set; relevant features are extracted from the integrated data set, which can be used to predict financial conditions. Then, the data set is input into an artificial intelligence algorithm model, such as a machine learning or deep learning model, and trained to obtain a prediction model; the trained prediction model is deployed in the financial system, and the company's financial status is predicted and analyzed by using the prediction model through real-time acquisition of financial information data, providing decision support and risk management.

[0069] The effects of the above technical solution are: by encrypting the financial information data, the security of the data during transmission and storage is ensured and data leakage is prevented; the encrypted financial information data is transmitted to the blockchain network, and each node can verify and record the data transmission and processing process to ensure the traceability of the data and prevent the data from being tampered with; the processed financial information data is stored on the alliance chain node, which has high credibility and security to ensure the integrity and confidentiality of the data; each node in the alliance chain calculates and processes the financial information data, and the processing result data is collected and integrated to form a comprehensive data set, which provides a basis for the subsequent prediction model training; through feature extraction and artificial intelligence algorithm model training, the accuracy of financial status prediction can be effectively improved. The trained prediction model is deployed in the financial system, and by obtaining financial information data in real time, the real-time prediction and analysis of the company's financial status can be achieved.

[0070] In one embodiment of the present invention, the method of obtaining financial information data from a financial system, encrypting the financial information data, and compressing the encrypted financial information data and transmitting it to a blockchain network includes:

[0071] Export the original financial information data that needs to be transmitted from the financial system and back up the exported financial information data;

[0072] Encrypt the financial information data using an AES encryption algorithm, and compress the encrypted financial information data using a compression algorithm (e.g., ZIP and LZMA);

[0073] Balancing the compression rate and compression speed of the compressed financial information data, and after the compression is completed, performing an integrity check on the data;

[0074] Split the integrity-checked data into multiple data slices, determine the optimal slice size and number based on network bandwidth and latency, and assign a unique identifier to each slice;

[0075] Use multi-channel transmission protocols and transmit data slices to the blockchain network through secure network channels (such as SSL and TLS), dynamically adjust transmission parameters according to network conditions (such as transmission rate, retry interval, etc.), and use breakpoint resume function to ensure that transmission can be continued from the interruption point when the network is interrupted. Set up a retry mechanism to automatically try to retransmit in the event of network failure or transmission failure.

[0076] The working principle of the above technical solution is as follows: first, the original financial information data to be transmitted is exported from the financial system, and the exported data is backed up to ensure data security and integrity; the financial information data is encrypted using the AES encryption algorithm, and then the encrypted data is compressed using a compression algorithm (such as ZIP or LZMA). This can protect the confidentiality of the data and reduce the storage space and transmission time required for transmission. The ZIP algorithm is usually faster but the compression rate may not be the highest, while algorithms such as LZMA may provide a higher compression rate but a slower processing speed. The most suitable algorithm is selected according to the characteristics of the data and the transmission requirements. For example, for a large amount of repeated data or data with predictable patterns, it may be more appropriate to use an algorithm with a higher compression rate; while for real-time transmission or applications that are sensitive to data delay, it may be more important to use a faster algorithm; the compressed data is divided to determine the optimal size and number of shards, specifically including: first evaluate the available network bandwidth. Obtain relevant information by measuring the transmission speed of the network connection or consulting the network service provider. Assume that the measured network bandwidth is 100Mbps; Next, evaluate the network delay. Test the delay with the target server or node by using the ping command or other network testing tools. Assume that the average delay measured is 50 milliseconds; calculate the optimal shard size based on the network bandwidth and delay. The bandwidth delay product (BDP) formula can be used. BDP is equal to bandwidth multiplied by delay. In this example, BDP = 100Mbps*50ms = 5Mb; based on the calculated BDP, the optimal shard size can be determined. Usually, the shard size is set to a value slightly smaller than the BDP to ensure that there will be no congestion or packet loss during network transmission. Assume that the optimal shard size is 4Mb; finally, determine the optimal number of shards based on the size of the data and the optimal shard size. It can be calculated by dividing the data size by the optimal shard size. For example, if the data size is 20Mb and the optimal shard size is 4Mb, the optimal number of shards is 20Mb / 4Mb = 5 shards. Assign a unique identifier to each shard for identification and management during transmission. Balance the compression rate and compression speed of the compressed financial information data, and after compression, perform an integrity check on the data; and balance the compression rate and compression speed based on user needs and business goals. For example, if saving storage space is the primary goal and transmission time is not a critical factor, a higher compression rate can be selected. On the contrary, if real-time performance is critical, some compression rate may need to be sacrificed to ensure faster transmission speed; adopt a multi-channel transmission protocol and transmit data slices to the blockchain network through a secure network channel (such as SSL or TLS). Dynamically adjust transmission parameters according to network conditions, including transmission rate, retry interval, etc., to ensure stable data transmission; after the transmission is completed, perform an integrity check on the transmitted data to ensure that the data has not been tampered with or lost.At the same time, a retry mechanism is established to automatically attempt retransmission in the event of network failure or transmission failure to ensure data reliability and integrity.

[0077] The effect of the above technical solution is: the financial information data is encrypted using the AES encryption algorithm to ensure that the data will not be accessed and stolen by unauthorized personnel during transmission. At the same time, a secure network channel (such as SSL or TLS) is used to transmit the data to the blockchain network to further ensure the security of the data; the encrypted financial information data is compressed by a compression algorithm (such as ZIP or LZMA) to reduce the storage space and transmission time required for transmission. This can improve transmission efficiency, reduce transmission costs, and complete data transmission faster when the network bandwidth is limited; dynamically adjust transmission parameters according to network conditions, including transmission rate, retry interval, etc. For example, when the network bandwidth is low, the transmission rate can be reduced to avoid data loss; when the network delay is high, the retry interval can be increased to improve the transmission success rate. Through these optimization adjustments, the stability and reliability of data transmission can be ensured; after the transmission is completed, the integrity of the transmitted data is checked to ensure that the data has not been tampered with or lost. If the data is found to be incomplete or erroneous, it can be automatically retransmitted to ensure the integrity and accuracy of the data; in the case of network interruption or transmission failure, the data can be continued from the interruption point through the breakpoint resume function to avoid repeated transmission of the transmitted data. This can save transmission time and bandwidth resources and improve transmission efficiency.

[0078] In one embodiment of the present invention, the blockchain node processes the financial information data accordingly after receiving it, and stores the processed financial information data on the alliance chain node, and the alliance chain node performs calculation processing on the data; including:

[0079] After receiving the financial information data, the blockchain node decrypts the financial information data using a decryption algorithm, and decompresses the decrypted financial information data; restoring the data to a readable mode;

[0080] The decrypted and decompressed financial information data is stored in the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of the financial information data; the financial information type includes transaction data, tax data and audit data.

[0081] Dividing the financial information data stored in different sub-nodes into multiple data blocks, and processing the data blocks by a parallel processing algorithm;

[0082] Monitor the computing resource load of each sub-node in real time during the processing process, use load balancing algorithm and resource scheduler to adjust the computing resources between each sub-node in real time;

[0083] The resource load calculation formula is:

[0084]

[0085] Where F represents the resource load, N represents the number of child nodes, and W i Represents the computing resources of a child node.

[0086] After the processing is completed, the processing result data of each child node is obtained.

[0087] For some time-consuming computing tasks, asynchronous processing is used; these tasks are placed in a message queue and processed asynchronously by background threads or other nodes.

[0088] The working principle of the above technical solution is as follows: after receiving the encrypted financial information data, the blockchain node uses the corresponding decryption algorithm to decrypt the data and restore it to a readable mode. The decrypted data may be large. In order to reduce storage space and transmission time, the blockchain node decompresses the data and restores the data to its original size; the decrypted and decompressed financial information data is stored on the alliance chain node. According to the type of financial information data (transaction data, tax data, audit data, etc.), the data will be stored in different sub-nodes; the alliance chain node uses a parallel processing algorithm to process the financial information data stored in different sub-nodes. This can improve processing efficiency while ensuring data consistency; during the processing process, the computing resource load of each sub-node is monitored in real time. Through the load balancing algorithm and resource scheduler, computing resources are dynamically allocated so that the computing resources between each sub-node can be reasonably utilized, thereby improving the overall processing efficiency. After the processing is completed, each sub-node returns the processing result data to the alliance chain node. These processing results contain the calculated and processed financial information data. For computing tasks that take a long time, asynchronous processing is adopted. These tasks will be placed in the message queue and processed asynchronously by background threads or other nodes. Whether a task is a time-consuming task is determined based on a preset time threshold.

[0089] The effect of the above technical solution is: by using the decryption algorithm to decrypt the financial information data, it is ensured that only authorized nodes can access and process the data. At the same time, the encryption transmission and storage technology is adopted to protect the security of the data during the transmission and storage process, and prevent data leakage and tampering; through the decompression operation, the financial information data is restored to a readable mode. According to the type of financial information, the alliance chain node stores the data in different sub-nodes to enhance the organization and management capabilities of the data. This makes it convenient for users to query and analyze the data and ensure the integrity and consistency of the data; the parallel processing algorithm and the load balancing algorithm are adopted to process multiple data blocks at the same time to improve the processing efficiency. The computing resource load of each sub-node is monitored in real time, and the allocation of computing resources is dynamically adjusted through the resource scheduler, so that the resources are reasonably utilized and the overall processing performance of the system is improved; for the computing tasks that take a long time, the asynchronous processing method is adopted to put the tasks into the message queue, and the background thread or other nodes are asynchronously processed. This can reduce the response time of the system, improve the concurrent processing capability of tasks, and reduce the occupation of system resources; the technical solution is based on blockchain and alliance chain technology and has high scalability. New nodes can be added dynamically according to actual needs to cope with the growing amount of data and user requests. At the same time, data is stored in different sub-nodes according to the data type, making data management more flexible and controllable. The above formula can achieve balanced distribution of resource load, avoid overload or idleness of a certain node, and improve the computing efficiency and performance of the overall system. Through real-time monitoring and dynamic adjustment, resources can be scheduled according to actual conditions so that the load can be more evenly distributed on each node. Such a design helps to improve the efficiency and reliability of blockchain nodes in processing financial information data. At the same time, the first part of the formula can better reflect the difference in computing volume between different nodes by weighted calculation of the difference between the computing resources and the average load of each sub-node. This helps to more fairly distribute computing tasks during resource scheduling and avoid overload or idleness of a certain node; the second part of the formula can comprehensively consider the average load of the entire system by calculating the weighted average of the average load of all nodes. This helps to achieve load balancing, avoid overload of a certain node, and improve the overall computing efficiency and performance; the calculation of the formula is based on real-time monitoring of the computing resource load of each sub-node during processing. The resource scheduler can adjust the computing resources between each sub-node in real time, and can make dynamic adjustments based on the actual load situation, so that each node can handle the load more evenly, improving the stability and performance of the overall system.

[0090] In one embodiment of the present invention, the processing result data after calculation and processing by each node in the alliance chain is collected, and the processing result data is integrated, and feature extraction is performed on the integrated processing result data to form a data set; the data set is input into an artificial intelligence algorithm model to obtain a prediction model; including:

[0091] Collect the processing result data after calculation from each node in the alliance chain, and send the result data to the federated learning server, and integrate the result data through the federated learning server;

[0092] Feature extraction is performed on the integrated data, and feature variable selection is performed, the extracted features are combined to form a data set, the data set is preprocessed, and the preprocessed data set is divided into a training set, a test set, and a validation set; the division ratio is 7:2:1.

[0093] Select an AI algorithm model based on the problem type and data characteristics, initialize the model parameters, send the data set to each node through the federated learning server, train the model with the training set, test the trained model with the test set, and verify the performance and generalization ability of the model with the validation set;

[0094] After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters.

[0095] After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0096] The working principle of the above technical solution is as follows: each node in the alliance chain will process data and generate processing result data. These processing result data will be collected and sent to the federated learning server, and the result data will be integrated through the federated learning server. Data privacy can be protected because the original data does not need to leave each node, and only the processing results are sent to the federated learning server; the integrated data will be feature extracted and feature variables will be selected to form a data set for training artificial intelligence algorithm models. Then the data set is preprocessed, including data cleaning, normalization and other operations to ensure the quality and availability of the data; the preprocessed data set will be divided into training set, test set and validation set according to the proportion. Then, according to the problem type and data characteristics, an appropriate artificial intelligence algorithm model is selected and the model parameters are initialized; the data set is sent to each node, each node uses the training set to train the model, and then uses the test set to test the trained model, and finally uses the validation set to verify the performance and generalization ability of the model. After the verification is completed, the local model parameters are sent to the federated learning server; after the federated learning server receives the model parameters of each node, it executes the federated average algorithm to aggregate the model parameters to obtain the global model parameters. Then, after receiving the global model parameters, each node updates the local model parameters and repeats the above steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0097] Specifically, in federated learning, the local data of each node cannot be shared directly, but the data is aggregated by training their own models and uploading the model parameters to the federated learning server to achieve joint training. Therefore, the data set division in the above steps is to ensure that the same data set is used in the model training process of each node, thereby ensuring the fairness and comparability of the federated learning model. The data set division is performed on the federated learning server, and then the divided data set is sent to each node. Each node has its own data locally, but only uses the data set provided by the federated learning server for model training. The model parameters trained by each node will be uploaded to the federated learning server for aggregation to obtain the global model, and the global model parameters will be sent to each node to update the local model parameters. This process is iterated multiple times until the preset stop condition is reached.

[0098] The effects of the above technical solution are as follows: Federated learning allows the original data to be processed locally at each node, and only the processing results are sent to the federated learning server for integration, effectively protecting data privacy; each node can share the results of model training, making full use of the data resources of distributed nodes and improving the training efficiency and performance of the model; the federated learning model update mechanism can achieve multi-party joint learning and model updating without exposing the original data, thereby improving the updating efficiency and accuracy of the model; the federated learning server performs feature extraction, feature selection and data preprocessing on the integrated data, effectively improving the quality and availability of the data set; the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters. After receiving the global model parameters, each node updates the local model parameters, repeats the iterative training process, and finally obtains the optimized prediction model.

[0099] In one embodiment of the present invention, the prediction model is deployed into a financial system, and the company's financial status is predicted by acquiring financial information data in real time; including:

[0100] Obtaining the required financial information data from the financial system in real time, preprocessing the financial information data, and predicting the preprocessed real-time data through a prediction model to obtain a prediction result of the company's financial status; the financial information data includes the company's balance sheet, cash flow statement, and income statement;

[0101] The real-time prediction results are displayed in a visual way, and the real-time performance of the prediction model is monitored and optimized and adjusted as needed.

[0102] The working principle of the above technical solution is: obtain the required financial information data from the financial system in real time, including the company's balance sheet, cash flow statement, and income statement; preprocess the obtained financial information data, including data cleaning, data conversion, feature extraction and other steps to ensure the accuracy and consistency of the data; input the preprocessed real-time data into the prediction model, and predict through the model to obtain the prediction results of the company's financial status. The prediction model can be built based on machine learning or deep learning algorithms, and the appropriate model can be selected according to the specific situation; the prediction results are displayed in a visual way, such as generating charts, reports or dashboards, so that users can intuitively understand the prediction of the company's financial status; monitor the real-time performance of the prediction model, including indicators such as model accuracy, stability and response time. Through real-time monitoring, it is possible to promptly detect model performance degradation or abnormalities, and take corresponding optimization and adjustment measures.

[0103] The effects of the above technical solutions are as follows: by obtaining financial information data from the financial system in real time, the company's latest financial status can be understood in a timely manner to avoid making wrong decisions due to data lags; by preprocessing real-time data and applying prediction models, the company's financial status can be accurately predicted, helping enterprises to promptly discover potential risks or opportunities and take corresponding measures; displaying the prediction results in a visual manner can intuitively present the company's financial status, enabling users to better understand and analyze data, thereby better making decisions and strategies; through real-time performance monitoring, performance problems of the prediction model can be discovered in a timely manner, such as decreased accuracy or prolonged response time, so that timely optimization and adjustment can be made to improve the stability and accuracy of the prediction model; by obtaining financial information in real time and accurately predicting the company's financial status, enterprises can make decisions efficiently, reduce decision-making risks, and improve the precision and accuracy of decisions.

[0104] One embodiment of the present invention, as Figure 2 As shown, a financial information processing system based on artificial intelligence, the system comprises:

[0105] Data transmission module: obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network;

[0106] Data processing module: after receiving the financial information data, the blockchain node processes it accordingly, and stores the processed financial information data on the alliance chain node, and performs calculation processing on the data through the alliance chain node; the corresponding processing includes decryption and decompression;

[0107] Model acquisition module: The model collects the processing result data after calculation and processing by each node in the alliance chain, integrates the processing result data, extracts features from the integrated processing result data, and forms a data set; the data set is input into the artificial intelligence algorithm model to obtain a prediction model;

[0108] Real-time prediction module: deploy the prediction model into the financial system and predict the company's financial status by acquiring financial information data in real time.

[0109] The working principle of the above technical solution is as follows: first, the system obtains financial information data from the financial system, and then encrypts the data to ensure data security and privacy protection; the encrypted financial information data is compressed to reduce the bandwidth and storage space required for data transmission. Then, the compressed data is transmitted to the blockchain network to ensure the traceability and tamper-proofness of the data; the blockchain node that receives the financial information data decrypts and decompresses it accordingly to restore the original data. The processed financial information data is stored on the alliance chain node, which has high credibility and security; each node in the alliance chain performs calculations on the processed financial information data, such as statistics and analysis. The processing result data is collected and integrated to form a comprehensive data set; relevant features are extracted from the integrated data set, which can be used to predict financial conditions. Then, the data set is input into an artificial intelligence algorithm model, such as a machine learning or deep learning model, and trained to obtain a prediction model; the trained prediction model is deployed in the financial system, and the company's financial status is predicted and analyzed by using the prediction model through real-time acquisition of financial information data, providing decision support and risk management.

[0110] The effects of the above technical solution are: by encrypting the financial information data, the security of the data during transmission and storage is ensured and data leakage is prevented; the encrypted financial information data is transmitted to the blockchain network, and each node can verify and record the data transmission and processing process to ensure the traceability of the data and prevent the data from being tampered with; the processed financial information data is stored on the alliance chain node, which has high credibility and security to ensure the integrity and confidentiality of the data; each node in the alliance chain calculates and processes the financial information data, and the processing result data is collected and integrated to form a comprehensive data set, which provides a basis for the subsequent prediction model training; through feature extraction and artificial intelligence algorithm model training, the accuracy of financial status prediction can be effectively improved. The trained prediction model is deployed in the financial system, and by obtaining financial information data in real time, the real-time prediction and analysis of the company's financial status can be achieved.

[0111] In one embodiment of the present invention, the data transmission module includes:

[0112] Information acquisition module: export the original financial information data to be transmitted from the financial system and back up the exported financial information data;

[0113] Data compression module: encrypts the financial information data using the AES encryption algorithm, and compresses the encrypted financial information data using a compression algorithm (e.g., ZIP and LZMA);

[0114] Data checking module: balances the compression rate and compression speed of the compressed financial information data, and performs integrity check on the data after compression;

[0115] Identifier allocation module: splits the integrity-checked data into multiple data slices, determines the optimal slice size and number based on network bandwidth and latency, and assigns a unique identifier to each slice;

[0116] Multi-channel transmission module: Use multi-channel transmission protocols and transmit data slices to the blockchain network through secure network channels (such as SSL and TLS), and dynamically adjust transmission parameters according to network conditions (such as transmission rate, retry interval, etc.), and use breakpoint resume function to ensure that transmission can be continued from the interruption point when the network is interrupted. Set up a retry mechanism to automatically try to retransmit in the event of network failure or transmission failure.

[0117] The working principle of the above technical solution is as follows: first, the original financial information data to be transmitted is exported from the financial system, and the exported data is backed up to ensure data security and integrity; the financial information data is encrypted using the AES encryption algorithm, and then the encrypted data is compressed using a compression algorithm (such as ZIP or LZMA). This can protect the confidentiality of the data and reduce the storage space and transmission time required for transmission. The ZIP algorithm is usually faster but the compression rate may not be the highest, while algorithms such as LZMA may provide a higher compression rate but a slower processing speed. The most suitable algorithm is selected according to the characteristics of the data and the transmission requirements. For example, for a large amount of repeated data or data with predictable patterns, it may be more appropriate to use an algorithm with a higher compression rate; while for real-time transmission or applications that are sensitive to data delays, it may be more important to use a faster algorithm; the compressed data is segmented to determine the optimal shard size and number, and a unique identifier is assigned to each shard for identification and management during transmission. The compression rate and compression speed of the compressed financial information data are balanced, and after the compression is completed, the data is checked for integrity; and the compression rate and compression speed are balanced based on user needs and business goals. For example, if saving storage space is the primary goal and transmission time is not a critical factor, a higher compression rate can be selected. On the contrary, if real-time performance is critical, some compression rates may need to be sacrificed to ensure faster transmission speeds; adopt a multi-channel transmission protocol and transmit data slices to the blockchain network through a secure network channel (such as SSL or TLS). Dynamically adjust transmission parameters according to network conditions, including transmission rate, retry interval, etc., to ensure stable data transmission; after the transmission is completed, perform an integrity check on the transmitted data to ensure that the data has not been tampered with or lost. At the same time, a retry mechanism is established to automatically attempt to retransmit in the event of a network failure or transmission failure to ensure data reliability and integrity.

[0118] The effect of the above technical solution is: the financial information data is encrypted using the AES encryption algorithm to ensure that the data will not be accessed and stolen by unauthorized personnel during transmission. At the same time, a secure network channel (such as SSL or TLS) is used to transmit the data to the blockchain network to further ensure the security of the data; the encrypted financial information data is compressed by a compression algorithm (such as ZIP or LZMA) to reduce the storage space and transmission time required for transmission. This can improve transmission efficiency, reduce transmission costs, and complete data transmission faster when the network bandwidth is limited; dynamically adjust transmission parameters according to network conditions, including transmission rate, retry interval, etc. For example, when the network bandwidth is low, the transmission rate can be reduced to avoid data loss; when the network delay is high, the retry interval can be increased to improve the transmission success rate. Through these optimization adjustments, the stability and reliability of data transmission can be ensured; after the transmission is completed, the integrity of the transmitted data is checked to ensure that the data has not been tampered with or lost. If the data is found to be incomplete or erroneous, it can be automatically retransmitted to ensure the integrity and accuracy of the data; in the case of network interruption or transmission failure, the data can be continued from the interruption point through the breakpoint resume function to avoid repeated transmission of the transmitted data. This can save transmission time and bandwidth resources and improve transmission efficiency.

[0119] In one embodiment of the present invention, the data processing module includes:

[0120] Data decryption module: After receiving the financial information data, the blockchain node decrypts the financial information data through a decryption algorithm, and decompresses the decrypted financial information data; restoring the data to a readable mode;

[0121] Multi-node storage module: The decrypted and decompressed financial information data is stored on the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of financial information data; the financial information type includes transaction data, tax data and audit data.

[0122] Parallel processing module: divides the financial information data stored in different sub-nodes into multiple data blocks, and processes the data blocks through parallel processing algorithms;

[0123] Resource balancing module: monitors the computing resource load of each sub-node in real time, uses load balancing algorithm and adjusts the computing resources between each sub-node in real time through resource scheduler;

[0124] Data acquisition module: After processing, obtain the processing result data of each sub-node.

[0125] For some time-consuming computing tasks, asynchronous processing is used; these tasks are placed in the message queue and processed asynchronously by background threads or other nodes.

[0126] The working principle of the above technical solution is as follows: after receiving the encrypted financial information data, the blockchain node uses the corresponding decryption algorithm to decrypt the data and restore it to a readable mode. The decrypted data may be large. In order to reduce storage space and transmission time, the blockchain node decompresses the data and restores the data to its original size; the decrypted and decompressed financial information data is stored on the alliance chain node. According to the type of financial information data (transaction data, tax data, audit data, etc.), the data will be stored in different sub-nodes; the alliance chain node uses a parallel processing algorithm to process the financial information data stored in different sub-nodes. This can improve processing efficiency while ensuring data consistency; during the processing process, the computing resource load of each sub-node is monitored in real time. Through the load balancing algorithm and resource scheduler, computing resources are dynamically allocated so that the computing resources between each sub-node can be reasonably utilized, thereby improving the overall processing efficiency. After the processing is completed, each sub-node returns the processing result data to the alliance chain node. These processing results contain the calculated and processed financial information data. For computing tasks that take a long time, asynchronous processing is adopted. These tasks will be placed in the message queue and processed asynchronously by background threads or other nodes. Whether a task is a time-consuming task is determined based on a preset time threshold.

[0127] The effect of the above technical solution is: by using the decryption algorithm to decrypt the financial information data, it is ensured that only authorized nodes can access and process the data. At the same time, the encryption transmission and storage technology is adopted to protect the security of the data during the transmission and storage process, and prevent data leakage and tampering; through the decompression operation, the financial information data is restored to a readable mode. According to the type of financial information, the alliance chain node stores the data in different sub-nodes to enhance the data organization and management capabilities. This makes it convenient for users to query and analyze the data and ensure the integrity and consistency of the data; by adopting parallel processing algorithms and load balancing algorithms, multiple data blocks can be processed at the same time to improve processing efficiency. The computing resource load of each sub-node is monitored in real time, and the allocation of computing resources is dynamically adjusted through the resource scheduler, so that resources are reasonably utilized and the overall processing performance of the system is improved; for computing tasks that take a long time, an asynchronous processing method is adopted to put the task into the message queue, and the background thread or other nodes process it asynchronously. This can reduce the response time of the system, improve the concurrent processing capability of tasks, and reduce the occupation of system resources; the technical solution is based on blockchain and alliance chain technology and has high scalability. New nodes can be added dynamically according to actual needs to cope with the growing amount of data and user requests. At the same time, data is stored in different sub-nodes according to the data type, making data management more flexible and controllable.

[0128] In one embodiment of the present invention, the model acquisition module includes:

[0129] Data integration module: collects the processing result data after calculation from each node in the alliance chain, and sends the result data to the federated learning server, and integrates the result data through the federated learning server;

[0130] Feature extraction module: extract features from the integrated data, select feature variables, combine the extracted features to form a data set, preprocess the data set, and divide the preprocessed data set into a training set, a test set, and a validation set; the division ratio is 7:2:1.

[0131] Model selection module: selects an AI algorithm model based on the problem type and data characteristics, initializes the model parameters, sends the data set to each node through the federated learning server, trains the model with the training set, tests the trained model with the test set, and verifies the performance and generalization ability of the model with the validation set;

[0132] Parameter aggregation module: After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters.

[0133] Model optimization module: After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0134] The working principle of the above technical solution is as follows: each node in the alliance chain will process data and generate processing result data. These processing result data will be collected and sent to the federated learning server, and the result data will be integrated through the federated learning server. Data privacy can be protected because the original data does not need to leave each node, and only the processing results are sent to the federated learning server; the integrated data will be feature extracted and feature variables will be selected to form a data set for training artificial intelligence algorithm models. Then the data set is preprocessed, including data cleaning, normalization and other operations to ensure the quality and availability of the data; the preprocessed data set will be divided into training set, test set and validation set according to the proportion. Then, according to the problem type and data characteristics, an appropriate artificial intelligence algorithm model is selected and the model parameters are initialized; the data set is sent to each node, each node uses the training set to train the model, and then uses the test set to test the trained model, and finally uses the validation set to verify the performance and generalization ability of the model. After the verification is completed, the local model parameters are sent to the federated learning server; after the federated learning server receives the model parameters of each node, it executes the federated average algorithm to aggregate the model parameters to obtain the global model parameters. Then, after receiving the global model parameters, each node updates the local model parameters and repeats the above steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

[0135] Specifically, in federated learning, the local data of each node cannot be shared directly, but the data are aggregated by training their own models and uploading the model parameters to the federated learning server to achieve joint training. Therefore, the data set division in the above steps is to ensure that the same data set is used in the model training process of each node, thereby ensuring the fairness and comparability of the federated learning model. Specifically, the data set division is performed on the federated learning server, and then the divided data set is sent to each node. Each node has its own data locally, but only uses the data set provided by the federated learning server for model training. The model parameters trained by each node will be uploaded to the federated learning server for aggregation to obtain the global model, and the global model parameters will be sent to each node to update the local model parameters. This process is iterated multiple times until the preset stop condition is reached.

[0136] The effects of the above technical solution are as follows: Federated learning allows the original data to be processed locally at each node, and only the processing results are sent to the federated learning server for integration, effectively protecting data privacy; each node can share the results of model training, making full use of the data resources of distributed nodes and improving the training efficiency and performance of the model; the federated learning model update mechanism can achieve multi-party joint learning and model updating without exposing the original data, thereby improving the updating efficiency and accuracy of the model; the federated learning server performs feature extraction, feature selection and data preprocessing on the integrated data, effectively improving the quality and availability of the data set; the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters. After receiving the global model parameters, each node updates the local model parameters, repeats the iterative training process, and finally obtains the optimized prediction model.

[0137] In one embodiment of the present invention, the real-time prediction module includes:

[0138] Status acquisition module: obtains the required financial information data from the financial system in real time, pre-processes the financial information data, and predicts the pre-processed real-time data through the prediction model to obtain the prediction result of the company's financial status; the financial information data includes the company's balance sheet, cash flow statement and income statement;

[0139] Visualization display module: Displays real-time prediction results in a visual way, monitors the real-time performance of the prediction model, and optimizes and adjusts it as needed.

[0140] The working principle of the above technical solution is: obtain the required financial information data from the financial system in real time, including the company's balance sheet, cash flow statement, and income statement; preprocess the obtained financial information data, including data cleaning, data conversion, feature extraction and other steps to ensure the accuracy and consistency of the data; input the preprocessed real-time data into the prediction model, and predict through the model to obtain the prediction results of the company's financial status. The prediction model can be built based on machine learning or deep learning algorithms, and the appropriate model can be selected according to the specific situation; the prediction results are displayed in a visual way, such as generating charts, reports or dashboards, so that users can intuitively understand the prediction of the company's financial status; monitor the real-time performance of the prediction model, including indicators such as model accuracy, stability and response time. Through real-time monitoring, it is possible to promptly detect model performance degradation or abnormalities, and take corresponding optimization and adjustment measures.

[0141] The effects of the above technical solutions are as follows: by obtaining financial information data from the financial system in real time, the company's latest financial status can be understood in a timely manner to avoid making wrong decisions due to data lags; by preprocessing real-time data and applying prediction models, the company's financial status can be accurately predicted, helping enterprises to promptly discover potential risks or opportunities and take corresponding measures; displaying the prediction results in a visual manner can intuitively present the company's financial status, enabling users to better understand and analyze data, thereby better making decisions and strategies; through real-time performance monitoring, performance problems of the prediction model can be discovered in a timely manner, such as decreased accuracy or prolonged response time, so that timely optimization and adjustment can be made to improve the stability and accuracy of the prediction model; by obtaining financial information in real time and accurately predicting the company's financial status, enterprises can make decisions efficiently, reduce decision-making risks, and improve the precision and accuracy of decisions.

[0142] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A financial information processing method based on artificial intelligence, characterized in that: The method comprises: Obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network; The blockchain network processes the financial information data accordingly after receiving it, and stores the processed financial information data on the alliance chain nodes, and performs calculations on the data through the alliance chain nodes; Collect the processing result data after calculation and processing by each node in the alliance chain, integrate the processing result data, extract features from the integrated processing result data, and form a data set; input the data set into the artificial intelligence algorithm model to obtain a prediction model; Deploy the forecasting model into the financial system and forecast the company's financial status by acquiring financial information data in real time; The blockchain network processes the financial information data accordingly after receiving it, and stores the processed financial information data on the alliance chain node, and performs calculation processing on the data through the alliance chain node; including: After receiving the financial information data, the blockchain network decrypts the financial information data using a decryption algorithm and decompresses the decrypted financial information data; The decrypted and decompressed financial information data is stored in the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of the financial information data; Dividing the financial information data stored in different sub-nodes into multiple data blocks, and processing the data blocks by a parallel processing algorithm; Monitor the computing resource load of each sub-node in real time during the processing process, use load balancing algorithm and resource scheduler to adjust the computing resources between each sub-node in real time; After the processing is completed, the processing result data of each child node is obtained; The processing result data after calculation and processing by each node in the alliance chain are collected, and the processing result data are integrated, and the feature extraction of the integrated processing result data is performed to form a data set; the data set is input into the artificial intelligence algorithm model to obtain a prediction model; including: Collect the processing result data after calculation from each node in the alliance chain, and send the result data to the federated learning server, and integrate the result data through the federated learning server; Extracting features from the integrated data and selecting feature variables, combining the extracted features to form a data set, preprocessing the data set, and dividing the preprocessed data set into a training set, a test set, and a validation set; Select an AI algorithm model based on the problem type and data characteristics, initialize the model parameters, send the data set to each node through the federated learning server, train the model with the training set, test the trained model with the test set, and verify the performance and generalization ability of the model with the validation set; After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters. After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

2. The method for processing financial information based on artificial intelligence according to claim 1, characterized in that: The method of obtaining financial information data from a financial system, encrypting the financial information data, and compressing the encrypted financial information data and transmitting it to a blockchain network includes: Export the original financial information data that needs to be transmitted from the financial system and back up the exported financial information data; Encrypt the financial information data using the AES encryption algorithm, and compress the encrypted financial information data using the compression algorithm; Balancing the compression rate and compression speed of the compressed financial information data, and after the compression is completed, performing an integrity check on the data; Split the integrity-checked data into multiple data slices, determine the optimal slice size and number based on network bandwidth and latency, and assign a unique identifier to each slice; A multi-channel transmission protocol is used to transmit data slices to the blockchain network through a secure network channel, and the transmission parameters are dynamically adjusted according to the network conditions.

3. The method for processing financial information based on artificial intelligence according to claim 1, characterized in that: The prediction model is deployed into the financial system, and the company's financial status is predicted by acquiring financial information data in real time; including: Obtain the required financial information data from the financial system in real time, pre-process the financial information data, and predict the pre-processed real-time data through the prediction model to obtain the prediction results of the company's financial status; The real-time prediction results are displayed in a visual way, and the real-time performance of the prediction model is monitored and optimized and adjusted as needed.

4. A financial information processing system based on artificial intelligence, characterized in that: The system comprises: Data transmission module: obtain financial information data from the financial system, encrypt the financial information data, compress the encrypted financial information data and transmit it to the blockchain network; Data processing module: after receiving the financial information data, the blockchain network processes it accordingly, stores the processed financial information data on the alliance chain nodes, and performs calculations on the data through the alliance chain nodes; Model acquisition module: The model collects the processing result data after calculation and processing by each node in the alliance chain, integrates the processing result data, extracts features from the integrated processing result data, and forms a data set; the data set is input into the artificial intelligence algorithm model to obtain a prediction model; Real-time prediction module: deploys the prediction model into the financial system and predicts the company's financial status by acquiring financial information data in real time; The data processing module comprises: Data decryption module: After receiving the financial information data, the blockchain network decrypts the financial information data through a decryption algorithm and decompresses the decrypted financial information data; Multi-node storage module: storing the decrypted and decompressed financial information data on the alliance chain node. After receiving the financial information data, the alliance chain node stores the financial information data in different sub-nodes according to the type of the financial information data; Parallel processing module: divides the financial information data stored in different sub-nodes into multiple data blocks, and processes the data blocks through parallel processing algorithms; Resource balancing module: monitors the computing resource load of each sub-node in real time, uses load balancing algorithm and adjusts the computing resources between each sub-node in real time through resource scheduler; Data acquisition module: After processing, obtain the processing result data of each sub-node; The model acquisition module includes: Data integration module: collects the processing result data after calculation from each node in the alliance chain, and sends the result data to the federated learning server, and integrates the result data through the federated learning server; Feature extraction module: extract features from the integrated data, select feature variables, combine the extracted features to form a data set, preprocess the data set, and divide the preprocessed data set into a training set, a test set, and a validation set; Model selection module: selects an AI algorithm model based on the problem type and data characteristics, initializes the model parameters, sends the data set to each node through the federated learning server, trains the model with the training set, tests the trained model with the test set, and verifies the performance and generalization ability of the model with the validation set; Parameter aggregation module: After verification, the local model parameters are sent to the federated learning server. After receiving the model parameters of each node, the federated learning server executes the federated averaging algorithm to aggregate the model parameters to obtain the global model parameters. Model optimization module: After receiving the global model parameters, each node updates the local model parameters and repeats the above two steps until the preset training rounds or convergence conditions are reached, and finally an optimized prediction model is obtained.

5. The financial information processing system based on artificial intelligence according to claim 4 is characterized in that: The data transmission module comprises: Information acquisition module: export the original financial information data to be transmitted from the financial system and back up the exported financial information data; Data compression module: encrypts financial information data using the AES encryption algorithm, and compresses the encrypted financial information data using the compression algorithm; Data checking module: balances the compression rate and compression speed of the compressed financial information data, and performs integrity check on the data after compression; Identifier allocation module: splits the integrity-checked data into multiple data slices, determines the optimal slice size and number based on network bandwidth and latency, and assigns a unique identifier to each slice; Multi-channel transmission module: It uses a multi-channel transmission protocol to transmit data slices to the blockchain network through a secure network channel, and dynamically adjusts transmission parameters according to network conditions.

6. The financial information processing system based on artificial intelligence according to claim 4, characterized in that: The real-time prediction module comprises: Status acquisition module: obtains the required financial information data from the financial system in real time, pre-processes the financial information data, predicts the pre-processed real-time data through the prediction model, and obtains the prediction result of the company's financial status; Visualization display module: Displays real-time prediction results in a visual way, monitors the real-time performance of the prediction model, and optimizes and adjusts it as needed.

Images