Financial management data processing method and system

Abstract

The invention provides a financial management data processing method and system, and relates to the field of financial data processing, and the method comprises the steps: traversing elements in a category data set, and dividing the category data set into at least one representation data chain according to the feature similarity between the elements; calculating data granularity of data items in the representation data chain, and obtaining a mother chain and sub-chain group according to the data granularity; performing hierarchical storage on the mother chain and the sub-chain, constructing an associated index map, and calculating a shortest query path from a mother chain layer to a sub-chain layer; and obtaining the user activeness of each access permission level corresponding to the shortest query path, and comparing the user activeness with the user activeness of each access permission level at the current moment to obtain a target query path at the current moment. According to the method, the activity degree of the user is monitored and the query path is dynamically adjusted, so that the change of user behaviors can be flexibly adapted, and the optimal data search path is found according to the current user request condition.

Description

Technical Field

[0001] The present invention relates to the field of financial data processing, and in particular to a financial management data processing method and system. Background Art

[0002] With the rapid growth of data in the financial industry, data management and processing has become a major challenge for financial institutions. Traditional data processing methods often face problems such as insufficient data security, low query efficiency, and low data retrieval accuracy when processing huge amounts of financial data; especially in the financial field, there are extremely high requirements for data security and access control. At the same time, the characteristics of financial data include but are not limited to large volume, high dimensionality, and timeliness, which puts higher demands on data processing and retrieval capabilities.

[0003] At present, most data processing methods use static data storage and query methods. This method often ignores the dynamic changes in user access patterns, resulting in long system response time and poor user experience in high-frequency access periods or complex query scenarios, and it is difficult to adapt to the rapidly changing characteristics of financial management data.

[0004] In view of this, this application is hereby filed. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a financial management data processing method and system, which can optimize the storage structure of financial management data and improve query efficiency and accuracy by dynamically adjusting the query path.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows: In a first aspect, a method for processing financial management data comprises: Presetting the user access permission level of the financial data, and classifying the financial data to obtain at least one category data set; Traversing the elements in the category data set, and dividing the category data set into at least one representation data chain according to feature similarities between the elements; Calculating the data granularity of the data items in the data chain, and obtaining a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; The parent chain and the child chain are stored in layers, an associated index graph is constructed, and the shortest query path from the parent chain layer to the child chain layer is calculated; The user activity of each access permission level corresponding to the shortest query path is obtained, and the target query path at the current moment is obtained by comparing it with the user activity of each access permission level at the current moment.

[0007] Furthermore, the user access permission level of the financial data is preset, and the financial data is classified and processed to obtain at least one category data set, including: Extract categorical features from financial data and preprocess the categorical features; Construct the corresponding feature vector for each data according to the classification features; Each data is clustered using the similarity between feature vectors through a clustering algorithm, and a category label is assigned to each data according to the clustering result to obtain at least one category data set.

[0008] Further, traversing the elements in the category data set, dividing the category data set into at least one representation data chain according to the feature similarity between the elements, including: Obtain the feature vector similarity and time series similarity between elements, and weightedly combine the feature vector similarity and time series similarity to obtain the comprehensive similarity between elements; Initialize an empty data link list and select an unassigned element as the starting point of the current data link; Traverse each unassigned element and calculate its comprehensive similarity with the tail point of the current data chain. If the comprehensive similarity is greater than the dynamic allocation threshold of the tail point of the current data chain, the data point is added to the data chain until the current data chain cannot add new points, and the characterization data chain is obtained.

[0009] Further, the data granularity of the data items in the data chain is calculated, and a parent chain and a child chain group are obtained according to the data granularity, including: Calculating the coupling coefficient between the data items according to the performance characteristics between the data items in the characterizing data chain; Construct a weighted graph to represent the data items in the data chain as nodes of the weighted graph, and use the inverse of the coupling coefficient between the data items as the weight of the edge between the nodes; Select any node as the starting node, set up an empty minimum spanning tree, and add the starting node to the minimum spanning tree; Among all the edges connecting nodes in the minimum spanning tree and nodes not in the minimum spanning tree, select the edge with the smallest weight, and add the edge and its corresponding node to the minimum spanning tree until all nodes are included in the minimum spanning tree, and the minimum spanning tree is used as the parent chain; For each non-parent chain data item, data items of the same level are grouped into a sub-chain according to the level information; Find the data item with the largest coupling coefficient between the sub-chain and the parent chain node, and form a connection pairing symbol between the sub-chain and the parent chain based on the data item and the parent chain node.

[0010] Furthermore, the coupling coefficient between the data items is calculated according to the performance characteristics between the data items in the characterizing data chain, including: Discretize each feature of the data item; For two data items and , calculate the data item and The joint occurrence count on all feature dimensions is divided by the total number of data points to get the joint probability ; For each data item , calculate the data item The number of occurrences in each feature dimension is divided by the total number of data points to get the marginal probability ; pass Calculated data items and The coupling coefficient between ,in, To characterize the data link.

[0011] Furthermore, for each data item that is not a parent chain, data items of the same level are grouped into a sub-chain according to the level information, including: Treat each non-parent chain data item as an independent cluster and calculate the distance between all cluster pairs; Perform multiple rounds of cluster merging operations. In each round, merge the two closest clusters into a new cluster and update the distance matrix between clusters. Constructing a hierarchical tree based on multiple rounds of cluster merging operations; Assign a level value to each data item according to the hierarchical tree, and group data items of the same level into a subchain.

[0012] Furthermore, the parent chain and the child chain are stored in layers, an associated index graph is constructed, and the shortest query path from the parent chain layer to the child chain layer is calculated, including: Generate a parent chain data packet and a child chain data packet for all data points in the parent chain and each child chain, store the parent chain data packet in the local storage node, store the child chain data packet in the cloud storage node, and create an index or identifier for each data packet; One or more connection points are defined for each subchain data packet, and the subchain data packets are sequentially linked according to the level value of the subchain to obtain a subchain data packet sequence; Create an associated index graph, calculate the distance from the parent chain data packet node to each node of the child chain data packet sequence based on the associated index graph, and select the minimum value among the distances to obtain the shortest distance; Obtain the connection point in the subchain data packet sequence corresponding to the shortest distance, and record it as the first connection point; The first connection point is used as the search starting point, and the search is performed at both ends of the sub-chain data packet sequence to obtain the shortest query path.

[0013] In a second aspect, a financial management data processing system includes: A classification module, used to preset the user access permission level of financial data and classify the financial data to obtain at least one category data set; A partitioning module, used for traversing the elements in the category data set, and partitioning the category data set into at least one representation data chain according to the feature similarity between the elements; A calculation module, used to calculate the data granularity of the data items in the data chain, and obtain a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; The storage module is used to store the parent chain and the child chain in layers, build an associated index graph, and calculate the shortest query path from the parent chain layer to the child chain layer; The query module is used to obtain the user activity of each access permission level corresponding to the shortest query path, and obtain the target query path at the current moment by comparing it with the user activity of each access permission level at the current moment.

[0014] According to a third aspect, a computing device includes: one or more processors; The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the above method.

[0015] In a fourth aspect, a computer-readable storage medium stores a program, and the program implements the above method when executed by a processor.

[0016] The above solution of the present invention includes at least the following beneficial effects: The above scheme of the present invention effectively strengthens data security and access control, can ensure the protection of sensitive data, and prevent unauthorized access; by monitoring user activity and dynamically adjusting the query path, it can flexibly adapt to changes in user behavior, find the optimal data search path according to the current user request situation, reduce unnecessary data traversal, further improve the efficiency of data query, and improve the user's query experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of a financial management data processing method provided by an embodiment of the present invention.

[0018] Figure 2 It is a module diagram of a financial management data processing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0020] like Figure 1 As shown, an embodiment of the present invention provides a financial management data processing method, the method comprising: Step 1, presetting the user access permission level of financial data, and classifying the financial data to obtain at least one category data set; Step 2, traversing the elements in the category data set, and dividing the category data set into at least one representation data chain according to the feature similarity between the elements; Step 3, calculating the data granularity of the data items in the data chain, and obtaining a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; Step 4: Store the parent chain and the child chain in layers, build an associated index graph, and calculate the shortest query path from the parent chain layer to the child chain layer; Step 5: Obtain the user activity of each access permission level corresponding to the shortest query path, and compare it with the user activity of each access permission level at the current moment to obtain the target query path at the current moment.

[0021] In the financial management data processing method described in the embodiment of the present invention, by setting the user access permission level and classifying the data, not only the data security in financial management is strengthened, but also the data retrievability is improved; by dividing the data into characterization data chains, the related data are logically connected, the correlation between the data is enhanced, and convenience is provided for subsequent joint query or data analysis; by granular division, the data is further refined into parent chains and child chains, and a multi-level data view is provided to support the data search requirements of different user access permission levels; by constructing an associated index graph, in the scenario of financial management, not only the storage structure of the data is optimized, but also the accessibility and retrieval efficiency of the data are improved; by calculating the shortest query path from the parent chain to the child chain layer, and dynamically adjusting the query path according to the comparison of the user activity corresponding to the shortest query path with the current activity, the query process is made more flexible, and can quickly adapt to changes in the current user behavior, and can find the optimal data search path according to the current user request situation, reduce unnecessary data traversal, and improve the query efficiency of the data.

[0022] In a preferred embodiment of the present invention, the above step 1, presetting the user access permission level of the financial data, and classifying the financial data to obtain at least one category data set, includes: Step 11, extracting classification features from the financial data and preprocessing the classification features; Step 12, constructing a corresponding feature vector for each data according to the classification feature; Step 13, clustering each data using the similarity between feature vectors through a clustering algorithm, and assigning a category label to each data according to the clustering result, to obtain at least one category data set.

[0023] In the above step 11, the classification features include but are not limited to product type, transaction volume, risk assessment, market performance indicators, etc. The preprocessing includes data cleaning to remove missing values ​​and outliers, and standardization to ensure that all data are compared on the same scale; for example, there is a large amount of loan application data, including the applicant's financial status, credit history, type and amount of loan applied for, etc. Classification features can be extracted from the loan application data, such as loan type (auto, housing, education, etc.), applicant's annual income, credit score, loan amount, historical default record, etc. These features are preprocessed, including processing missing values ​​(such as filling with median), outliers (such as identifying and processing through IQR method) and standardization (such as using Z-score standardization). Thus, each loan application is converted into a feature vector, which contains preprocessed data points such as loan type, annual income, credit score, etc. Clustering algorithms (such as K-means) are then used to cluster loan applications according to the similarity of feature vectors. Each cluster represents a class of loan applications with similar risk characteristics. According to the clustering results, a category label is assigned to each application, such as low-risk, medium-risk, and high-risk loans.

[0024] In the financial management data processing method described in the embodiment of the present invention, through data preprocessing, including data cleaning and standardization, noise and inconsistency in the data can be removed, the data quality is improved, and an accurate data basis is provided for subsequent analysis; the preprocessing step can ensure that all data are compared on the same scale, eliminating the deviations that may be introduced by different data sources and data types; grouping the data through a clustering algorithm can reveal the inherent structure and pattern of the data, and help understand the essential characteristics and potential associations of the data; assigning category labels to each data facilitates subsequent classification analysis, pattern recognition and decision support, and improves the accuracy and depth of data analysis.

[0025] In a preferred embodiment of the present invention, the above step 2, traversing the elements in the category data set and dividing the category data set into at least one representation data chain according to the feature similarity between the elements, includes: Step 21, obtaining the feature vector similarity and time series similarity between elements, and weightedly merging the feature vector similarity and time series similarity to obtain the comprehensive similarity between elements; Step 22, by formula: Computational Elements Dynamic allocation threshold ;in, The average of the initial similarity thresholds of all elements, used for normalization; the total number of elements in the N-category dataset; The number of nearest neighbors used to compute the local density; and They are the elements x i and its Nearest Neighbor In the The eigenvalue of the dimension; and are element x and its Nearest Neighbor No. The eigenvalue of the dimension; is the dimension of the feature; Step 23, initialize an empty data link list and select an unassigned element as the starting point of the current data link; Step 24, traverse each unallocated element, calculate its comprehensive similarity with the tail point of the current data chain, if the comprehensive similarity is greater than the dynamic allocation threshold of the tail point of the current data chain, then add the data point to the data chain, until the current data chain cannot add new points, and obtain the characterization data chain; Step 25, repeating the above steps 23 and 24 until all elements in the category data set are allocated.

[0026] In the financial management data processing method described in the embodiment of the present invention, the elements in the category data set are divided into characterization data chains, each data chain represents a group of elements with similar characteristics. This organization method makes the structure and relationship of the data clearer, improves the interpretability of the data, and helps to deeply understand the intrinsic characteristics of the data; by using the weighted combination of feature vector similarity and time series similarity, the multi-faceted similarities between elements can be comprehensively considered to ensure the comprehensiveness and accuracy of the similarity calculation. The calculation of the dynamic allocation threshold takes into account the local density of the data, making the construction of the data chain more refined and dynamic, and able to adapt to the distribution characteristics of the elements in the data set, thereby improving the accuracy of subsequent analysis. Data can be divided into data chains with intrinsic similarity, and the storage structure of the data can be optimized, so that data points with correlation are organized together, which not only saves storage space, but also improves the efficiency and accuracy of data retrieval. The constructed data chain provides a basis for complex data mining and pattern recognition tasks, such as dividing stocks into different data chains according to their behavior patterns. For example, some data chains may represent stocks with stable growth, while others may represent stocks with large fluctuations. First, the average daily price and trading volume of each stock are extracted as feature vectors, and the similarity of time series is considered. For example, dynamic time warping (DTW) is used to calculate the similarity between price time series. The feature vector similarity and time series similarity are weighted and combined to obtain the comprehensive similarity between each pair of stocks. Then, according to the provided formula, the dynamic allocation threshold of each stock is calculated, taking into account the local density and similarity of the feature values ​​between stocks, ensuring that the threshold can reflect the local relationship and overall distribution between stocks. Then, an unallocated stock is selected as the starting point of the current data chain, and an empty data chain list is initialized. Finally, the unallocated stocks are traversed and their comprehensive similarity with the stocks at the end of the current data chain is calculated. If the similarity exceeds the dynamic allocation threshold, the stock is added to the current data chain. Continue this process until no more stocks can be added to the current data chain. And until all stocks are assigned to a data chain, finally multiple data chains representing different behavior patterns are obtained.

[0027] In a preferred embodiment of the present invention, the above step 3, calculating the data granularity of the data items in the data chain, and obtaining a parent chain and a child chain group according to the data granularity, includes: Step 31, calculating the coupling coefficient between the data items according to the performance characteristics between the data items in the characterizing data chain; Step 32, construct a weighted graph to represent the data items in the data chain as nodes of the weighted graph, and use the inverse of the coupling coefficient between the data items as the weight of the edge between the nodes; Step 33, select any node as the starting node, set an empty minimum spanning tree, and add the starting node to the minimum spanning tree; Step 34, among all the edges connecting the nodes in the minimum spanning tree and the nodes not in the minimum spanning tree, select the edge with the smallest weight, and add the edge together with its corresponding node to the minimum spanning tree until all the nodes are included in the minimum spanning tree, and use the minimum spanning tree as the parent chain; Step 35, for each data item that is not a parent chain, group the data items at the same level into a child chain according to the level information; Step 36, find the data item with the largest coupling coefficient between the sub-chain and the parent chain node, and form a connection pairing symbol between the sub-chain and the parent chain based on the data item and the parent chain node.

[0028] In the financial management data processing method described in the embodiment of the present invention, by constructing a weighted graph and a minimum spanning tree, complex data relationships can be simplified into structures that are easy to manage and interpret, namely, parent chains and child chain groups. It is not only convenient for data storage and access, but also helps to maintain the relevance and hierarchy between data; using the minimum spanning tree as the parent chain ensures the efficient expression of the relationship between data. When retrieving or querying data, the parent chain provides an efficient access path, thereby reducing the time and computing resource consumption of data access. By determining the connection point between the child chain and the parent chain, a specific data subset can be quickly located and accessed. Calculating the coupling coefficient between data items and applying it to the construction of a weighted graph can quantify the relationship between data and provide a clear perspective for the interaction and dependence between data items. The construction of the child chain in this method is based on hierarchical information, which helps to reveal the hierarchical structure between data items. Through this structured method, the addition, update or deletion of data can be performed more efficiently. Especially in a dynamically changing data environment, the structure of the parent chain and the child chain facilitates the implementation of incremental updates and maintains data consistency.

[0029] In a preferred embodiment of the present invention, the above step 31, calculating the coupling coefficient between the data items according to the performance characteristics between the data items in the characterizing data chain, includes: Step 311, discretize each feature of the data item, for example, if the feature is continuous, it can be converted into discrete values ​​by a binning method; Step 312, for two data items and , calculate the data item and The joint occurrence count on all feature dimensions is divided by the total number of data points to get the joint probability ; Step 313, for each data item , calculate the data item The number of occurrences in each feature dimension is divided by the total number of data points to get the marginal probability ; Step 314, by Calculated data items and The coupling coefficient between ,in, To characterize the data link.

[0030] In the financial management data processing method described in the embodiment of the present invention, for example, to evaluate the relationship between different stocks in stock data, first select the characteristics of the stock, such as daily yield, trading volume, etc. If these characteristics are continuous, they can be converted into discrete values ​​by the binning method. For example, the daily yield is divided into three boxes of "high yield", "medium yield" and "low yield"; then for stocks A and B, their joint occurrence times in all feature dimensions (such as daily yield and trading volume) are calculated. For example, the number of days when stocks A and B are in the "high yield" state at the same time is calculated, and then divided by the total number of trading days to obtain the joint probability; then the number of occurrences of stocks A and B in each feature dimension is calculated, and then divided by the total number of trading days to obtain the marginal probability. For example, the proportion of the number of days when stock A is in the "high yield" state to the total number of trading days is calculated. Finally, the coupling coefficient between stocks A and B is calculated using the joint probability and the marginal probability. This coefficient can quantify the strength of the relationship between the two stocks.

[0031] By using the joint probability and the marginal probability to calculate the coupling coefficient, we can deeply understand the relationship between different data items, whether it is positive correlation, negative correlation or irrelevant. The discretization process in step 311 can convert continuous data into discrete values ​​that are easy to handle, reducing the complexity of the data while retaining the key information of the data.

[0032] In a preferred embodiment of the present invention, the above step 35, for each data item of a non-parent chain, groups data items of the same level into a sub-chain according to the level information, including: Step 351, treat each non-parent link data item as an independent cluster, and calculate the distance between all cluster pairs; wherein the distance can be calculated in the form of the shortest distance (single connection), the longest distance (full connection), the average distance (average connection), etc., which can be achieved through corresponding related algorithms.

[0033] Step 352, performing multiple rounds of cluster merging operations, in each round, finding the two clusters with the closest distance, merging them into a new cluster, and updating the distance matrix between the clusters; Step 353, construct a hierarchical tree based on multiple rounds of clustering and merging operations; wherein each merging operation corresponds to a node on the hierarchical tree, each data item is a leaf node of the tree at the beginning, and each merge generates a new internal node. The structure of the tree is determined according to the order and hierarchy of the clustering and merging, and the level of each data item in the tree is also determined.

[0034] Step 354, assign a level value to each data item according to the hierarchical tree, and group data items of the same level into a subchain; wherein each layer of the hierarchical tree represents a level, the root node represents the result of clustering merging, and is located at the top layer, and the leaf node represents a single data item, and is located at the bottom layer; a level value is assigned to each data item, which value can be the depth of the tree in which it is located, or the number of edges required to pass from the root node to the node.

[0035] In the financial management data processing method described in the embodiment of the present invention,.

[0036] In a preferred embodiment of the present invention, after step 35, the method further includes: step 355, mapping the user access permission level to the sub-chain so that the number of sub-chains matches the user access permission level. Specifically, through the formula Map each subchain i to the corresponding access permission level, where ; is a mathematical function. The "ceil(x)" function returns the smallest integer greater than or equal to the number "x". That is, "ceil" rounds "x" up to the nearest integer. For example, the result of ceil(1.2) is 2, the result of ceil(2.0) is 2, and the result of ceil(-3.4) is -3 (because -3 is the smallest integer greater than -3.4). Merge multiple subchains mapped in the same access permission level, update the subchain, and use the average level value as the updated level value of the updated subchain.

[0037] In the financial management data processing method described in the embodiment of the present invention, for example, assuming that a bank has a large amount of credit approval data, which needs to be managed according to the user's credit level (as the access permission level), the credit approval data is first organized into multiple sub-chains according to the characteristics. Each sub-chain represents a group of credit approval records with similar characteristics. Then, a formula is used to map each sub-chain to the corresponding access permission level. Assuming that there are 5 access permission levels in total, the bank evenly distributes these sub-chains to these 5 levels; in the same access permission level, there may be multiple sub-chains mapped to this level. The bank merges these sub-chains into a larger sub-chain to simplify data management. During the merging process, the hierarchical value of each atomic chain is taken into account, and the hierarchical value of the merged new sub-chain is the average of the atomic chain hierarchical values. Thus, bank employees can quickly access the corresponding level of credit approval data according to their access permission level. For example, a senior credit analyst can access high-risk credit records, which are in a higher access permission level. By layering and sub-chaining data, data can be managed and queried more efficiently, and the speed and accuracy of processing data search can be improved. At the same time, this approach also ensures the security of data access, ensuring that only appropriate personnel can access data of the corresponding level, preventing data leakage or unauthorized access.

[0038] In a preferred embodiment of the present invention, the above step 4, storing the parent chain and the child chain in layers, constructing an associated index graph, and calculating the shortest query path from the parent chain layer to the child chain layer, includes: Step 41, generate a parent chain data packet and a child chain data packet for all data points in the parent chain and each child chain, store the parent chain data packet in the local storage node, store the child chain data packet in the cloud storage node, and create an index or identifier for each data packet; wherein the data packet includes information and related metadata of all data points in the parent chain or child chain, such as data point ID, feature value, etc.; Step 42, defining one or more connection points for each subchain data packet, and sequentially linking the subchain data packets according to the level value of the subchain to obtain a subchain data packet sequence; Step 43, creating an association index graph, in which each node stores the index or identifier of the data packet, the type of the data packet, the storage location, and the information of the connection point; specifically, in the graph, the edge between the nodes represents the association relationship between the data packets, and for each pair of adjacent data packets in the sub-chain data packet sequence, an edge is established to represent the order relationship between them, and for the parent chain data packet and the sub-chain data packet, the connection pairing symbol between the parent chain and the sub-chain is recorded on the corresponding node; Step 44, calculate the distance from the parent chain data packet node to each node of the child chain data packet sequence according to the association index graph, and select the minimum value among the distances to obtain the shortest distance; Step 45, obtaining a connection point in the sub-chain data packet sequence corresponding to the shortest distance, recorded as the first connection point; Step 46, taking the first connection point as the search starting point, and searching at both ends of the sub-chain data packet sequence to obtain the shortest query path.

[0039] In the financial management data processing method described in the embodiment of the present invention, by dividing the data into a parent chain and a child chain, and storing them locally and in the cloud respectively, not only can the data access speed be optimized, but also the storage strategy can be dynamically adjusted according to the access frequency and importance of the data, thereby reducing the storage cost and improving the data access efficiency. An index or identifier is created for each data packet, and an associated index graph is constructed, so that the relationship between the data packets is clearer. When querying, the child chain and the parent chain can be quickly merged directly through the index, which greatly improves the query speed and efficiency. By defining a connection point for each child chain data packet and linking them sequentially according to the hierarchical value, it is convenient to store on the child chain layer and to facilitate subsequent data search. The shortest query path from the parent chain to the child chain is calculated to reduce the retrieval of irrelevant data and shorten the query time, so that reverse reasoning can be performed to determine the optimal saturated user activity of each permission level matched under the shortest query path, and the connection point under the shortest distance is selected as the query starting point, which further refines the query entry point, thereby establishing a data query channel from the parent chain layer to the child chain layer during the query process.

[0040] In a preferred embodiment of the present invention, the above step 5, obtaining the user activity of each access permission level corresponding to the shortest query path, and comparing it with the user activity of each access permission level at the current moment, to obtain the target query path at the current moment, includes: Step 51, obtaining the number of time steps of the subchain data packets corresponding to each access authority level under the shortest query path, recorded as the first time step set; Step 52, arranging the elements in the first time step set from low to high order to obtain the first user activity distribution corresponding to each access authority level; Step 53: Obtain the user activity distribution of each access permission level at the current moment, and use the formula Calculate the activity change for each permission level ,in, is the user activity at the current moment corresponding to permission level i, is the first user activity distribution corresponding to permission level i; Step 54, for each subchain data packet, by formula , calculate the comprehensive score of each subchain data packet ,in, and The coefficient used to balance the impact of two indicators; Step 55, select comprehensive rating The largest subchain data packet and its corresponding connection point set; Step 56: Select the element in the connection point set that is closest to the first connection point as the second connection point. Step 57, taking the second connection point as the search starting point, and searching at both ends of the sub-chain data packet sequence at the same time, to obtain the target query path at the current moment.

[0041] In the financial management data processing method described in the embodiment of the present invention, by obtaining the user activity of each access right level under the shortest query path and comparing it with the activity at the current moment, the change of user query behavior can be accurately captured. By calculating a comprehensive score for each sub-chain data packet, the method comprehensively considers the change of user activity and the current activity, thereby ensuring that the sub-chain data packet with the highest comprehensive score can correspond to the optimal query path entry point at this time, and then under the reference of the first connection point corresponding to the shortest distance, the connection point in the sub-chain data packet with the highest comprehensive score that is away from the first connection point is selected as the second connection point, which can further make the query path shortest, ensuring that the query process can more quickly approach the data that the current user is most concerned about, thereby improving the query response speed and accuracy; this method adjusts the real-time query path according to the user activity at each moment, so that the data retrieval process is more in line with the actual needs of the user, and improves the relevance and efficiency of the retrieval.

[0042] By presetting the user access rights level of financial data and performing detailed data classification processing, the present invention effectively strengthens the security and access control of data. It can ensure the protection of sensitive data and prevent unauthorized access, thereby achieving a high degree of security in the management process of financial data. By using feature similarity to characterize the division of data chains, and by calculating the data granularity to construct the strategy of parent chain and child chain group, the present invention significantly improves the retrieval efficiency of data. Through refined data organization and hierarchical management, the correlation between financial data is enhanced, which provides convenience for joint query or data analysis. This structured data processing method not only optimizes the storage structure of data, but also improves the accessibility and retrieval efficiency of data, so that users can retrieve relevant data more accurately and quickly as needed. By monitoring user activity and dynamically adjusting the query path, the present invention can flexibly adapt to changes in user behavior and find the optimal data search path according to the current user request situation. This dynamic adjustment mechanism reduces unnecessary data traversal, further improves the efficiency of data query, and improves the user's query experience. The present invention provides a multi-level data view and hierarchical storage strategy to support data search requirements of different user access rights levels, making data management more flexible. In addition, through the construction of clear data structures and index graphs, data addition, update or deletion operations are more efficient and easier to maintain, reducing the cost of data management.

[0043] In a second aspect, the present invention further provides a financial management data processing system, comprising: A classification module 10, used to preset a user access permission level for financial data and classify the financial data to obtain at least one category data set; A partitioning module 20, configured to traverse the elements in the category data set and partition the category data set into at least one representation data chain according to feature similarities between the elements; A calculation module 30, used to calculate the data granularity of the data items in the data chain, and obtain a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; Storage module 40, used to store parent chains and child chains in layers, build an associated index graph, and calculate the shortest query path from the parent chain layer to the child chain layer; The query module 50 is used to obtain the user activity of each access permission level corresponding to the shortest query path, and obtain the target query path at the current moment by comparing it with the user activity of each access permission level at the current moment.

[0044] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0045] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0046] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0047] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0048] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0049] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0050] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0051] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0052] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0053] In addition, it should be pointed out that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understandable that all or any steps or components of the method and apparatus of the present invention can be implemented in hardware, firmware, software or a combination thereof in any computing device (including processors, storage media, etc.) or a network of computing devices, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0054] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0055] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for processing financial management data, characterized in that: The method comprises: Presetting the user access permission level of the financial data, and classifying the financial data to obtain at least one category data set; Traversing the elements in the category data set, and dividing the category data set into at least one representation data chain according to feature similarities between the elements; Calculating the data granularity of the data items in the data chain, and obtaining a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; The parent chain and the child chain are stored in layers, an associated index graph is constructed, and the shortest query path from the parent chain layer to the child chain layer is calculated; The user activity of each access permission level corresponding to the shortest query path is obtained, and the target query path at the current moment is obtained by comparing it with the user activity of each access permission level at the current moment.

2. The financial management data processing method according to claim 1, characterized in that: Preset the user access permission level of financial data, and classify the financial data to obtain at least one category data set, including: Extract categorical features from financial data and preprocess the categorical features; Construct the corresponding feature vector for each data according to the classification features; Each data is clustered using the similarity between feature vectors through a clustering algorithm, and a category label is assigned to each data according to the clustering result to obtain at least one category data set.

3. The financial management data processing method according to claim 2, characterized in that: Traversing the elements in the category data set, and dividing the category data set into at least one representation data chain according to the feature similarity between the elements, including: Obtain the feature vector similarity and time series similarity between elements, and weightedly combine the feature vector similarity and time series similarity to obtain the comprehensive similarity between elements; Initialize an empty data link list and select an unassigned element as the starting point of the current data link; Traverse each unassigned element and calculate its comprehensive similarity with the tail point of the current data chain. If the comprehensive similarity is greater than the dynamic allocation threshold of the tail point of the current data chain, the data point is added to the data chain until the current data chain cannot add new points, and the characterization data chain is obtained.

4. The financial management data processing method according to claim 3, characterized in that: Calculating the data granularity of the data items in the characterizing data chain, and obtaining a parent chain and a child chain group according to the data granularity, including: Calculating the coupling coefficient between the data items according to the performance characteristics between the data items in the characterizing data chain; Construct a weighted graph to represent the data items in the data chain as nodes of the weighted graph, and use the inverse of the coupling coefficient between the data items as the weight of the edge between the nodes; Select any node as the starting node, set up an empty minimum spanning tree, and add the starting node to the minimum spanning tree; Among all the edges connecting nodes in the minimum spanning tree and nodes not in the minimum spanning tree, select the edge with the smallest weight, and add the edge and its corresponding node to the minimum spanning tree until all nodes are included in the minimum spanning tree, and the minimum spanning tree is used as the parent chain; For each non-parent chain data item, data items of the same level are grouped into a sub-chain according to the level information; Find the data item with the largest coupling coefficient between the sub-chain and the parent chain node, and form a connection pairing symbol between the sub-chain and the parent chain based on the data item and the parent chain node.

5. The financial management data processing method according to claim 4, characterized in that: Calculating the coupling coefficient between the data items according to the performance characteristics between the data items in the characterizing data chain includes: Discretize each feature of the data item; For two data items and , calculate the data item and The joint occurrence count on all feature dimensions is divided by the total number of data points to get the joint probability ; For each data item , calculate the data item The number of occurrences in each feature dimension is divided by the total number of data points to get the marginal probability ; pass Calculated data items and The coupling coefficient between ,in, To characterize the data link.

6. The financial management data processing method according to claim 5, characterized in that: For each non-parent chain data item, data items of the same level are grouped into a sub-chain according to the level information, including: Treat each non-parent chain data item as an independent cluster and calculate the distance between all cluster pairs; Perform multiple rounds of cluster merging operations. In each round, merge the two closest clusters into a new cluster and update the distance matrix between clusters. Constructing a hierarchical tree based on multiple rounds of cluster merging operations; Assign a level value to each data item according to the hierarchical tree, and group data items of the same level into a subchain.

7. The financial management data processing method according to claim 6, characterized in that: The parent chain and the child chain are stored in layers, an associated index graph is constructed, and the shortest query path from the parent chain layer to the child chain layer is calculated, including: Generate a parent chain data packet and a child chain data packet for all data points in the parent chain and each child chain, store the parent chain data packet in the local storage node, store the child chain data packet in the cloud storage node, and create an index or identifier for each data packet; One or more connection points are defined for each subchain data packet, and the subchain data packets are sequentially linked according to the level value of the subchain to obtain a subchain data packet sequence; Create an associated index graph, calculate the distance from the parent chain data packet node to each node of the child chain data packet sequence based on the associated index graph, and select the minimum value among the distances to obtain the shortest distance; Obtain the connection point in the subchain data packet sequence corresponding to the shortest distance, and record it as the first connection point; The first connection point is used as the search starting point, and the search is performed at both ends of the sub-chain data packet sequence to obtain the shortest query path.

8. A financial management data processing system, characterized in that: include: A classification module, used to preset the user access permission level of financial data and classify the financial data to obtain at least one category data set; A partitioning module, used for traversing the elements in the category data set, and partitioning the category data set into at least one representation data chain according to the feature similarity between the elements; A calculation module, used to calculate the data granularity of the data items in the data chain, and obtain a parent chain and a child chain group according to the data granularity, wherein the child chain group includes at least one child chain; The storage module is used to store the parent chain and the child chain in layers, build an associated index graph, and calculate the shortest query path from the parent chain layer to the child chain layer; The query module is used to obtain the user activity of each access permission level corresponding to the shortest query path, and obtain the target query path at the current moment by comparing it with the user activity of each access permission level at the current moment.

9. A computing device, characterized in that include: one or more processors; one or more processors; A storage device, used for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images