Data storage optimization method and platform for business integration

Abstract

The application provides a data storage optimization method and platform for business integration, relates to the technical field of data processing, and comprises the following steps: extracting the department architecture of a target enterprise to obtain a plurality of business function departments; calling business data to obtain a plurality of historical business data; performing business interaction analysis to construct and generate a data integration topology network; according to the data flow direction, the plurality of historical business data are divided into a plurality of groups of business data subsets; a plurality of business cooperation judgment models are constructed and mapped and synchronized to a plurality of business data uploading nodes; the data flow direction of dynamically uploaded business data is predicted, intelligent data routing is performed based on the prediction result, and data storage updating of K intermediates is completed. The application solves the technical problem that the data management and storage of different business departments in an enterprise are usually static, isolated, lack intelligent and dynamic optimization capabilities, and the overall business process efficiency is low.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a data storage optimization method and platform for business integration. BACKGROUND

[0002] In the era of big data, enterprises need to process a large amount of data, and often involve different data sources and multi-department cooperation. How to optimize data storage and management and improve the efficiency of data analysis is one of the core problems that need to be solved in the field of big data. Specifically, the data of different business departments is usually stored in independent systems, forming data islands, which makes it difficult to share and cooperate across departments, and the information circulation is not smooth, which affects the efficiency of the overall business process. The existing technology usually bases on static rules or manual configuration when dealing with cross-department data integration, lacks in-depth understanding and intelligent analysis capability of the internal relationship of data, which makes it difficult to cope with complex and variable business scenarios, and cannot make real-time intelligent judgment and decision, affecting the flexibility and accuracy of business cooperation. SUMMARY

[0003] The present application provides a data storage optimization method and platform for business integration, aiming to solve the technical problem that the data management and storage of different business departments in the existing technology are usually static and isolated, lacking intelligent and dynamic optimization capability, resulting in low efficiency of the overall business process.

[0004] The first aspect of the present application provides a data storage optimization method for business integration, the method comprising: extracting the department architecture of a target enterprise to obtain a plurality of business function departments; calling business data of the plurality of business function departments to obtain a plurality of historical business data; analyzing the business interaction of the plurality of business function departments based on the plurality of historical business data to construct a data integration topology network, wherein the data integration topology network comprises a plurality of business data upload nodes mapped to the plurality of business function departments and K intermediates, the K intermediates are used for pre-storing business integration data; according to the data flow direction of the plurality of historical business data, the plurality of historical business data are divided into a plurality of business data subsets, wherein each business data subset in the plurality of business data subsets comprises a local data subset and a plurality of cross-department data subsets; a plurality of business cooperation judgment models are constructed based on the plurality of business data subsets, and the plurality of business cooperation judgment models are mapped and synchronized to the plurality of business data upload nodes to complete the function enhancement of the data integration topology network; in the data integration topology network, the plurality of business data upload nodes predict the data flow direction of the business data dynamically uploaded by the plurality of business function departments through the plurality of business cooperation judgment models, and intelligently route the business data based on the prediction results to complete the data storage update of the K intermediates.

[0005] In a second aspect, the application discloses a data storage optimization platform for business integration, which is used for the data storage optimization method for business integration, and comprises a department architecture extraction module, a business data calling module, a business interaction analysis module, a historical business data segmentation module, and a topology network function enhancement module.

[0006] The one or more technical solutions provided in the application have at least the following technical effects or advantages:

[0007] By extracting the department architecture of the target enterprise and calling the historical business data of each department, the business flow within the enterprise can be comprehensively understood, and the interaction relationship and data dependency between departments can be effectively identified, laying a foundation for subsequent data integration; based on the historical business data, a data integration topology network is constructed, which effectively reflects the interaction relationship between each business function department within the enterprise, and through mapping multiple business data upload nodes and K intermediates in the network, the pre-storage and management of business data are realized, such topology structure not only optimizes the distribution of data storage, but also improves the speed and efficiency of data access, ensuring the efficient flow of data between departments; by analyzing the data flow of historical business data, the data is divided into multiple sets of business data subsets, and on this basis, multiple business collaboration judgment models are constructed, which can deeply analyze the explicit and implicit information of business data and identify the relevance between data, such modeling design enables the system to intelligently judge whether the business data needs to be uploaded or flowed between different departments, thereby enhancing the collaboration ability and decision support between departments; at the data upload node, the business collaboration judgment model is used to predict the flow direction of real-time uploaded data, and intelligent data routing is performed based on the prediction result, ensuring that the data can be timely and accurately transmitted to the appropriate intermediate for processing or storage, thereby optimizing the resource utilization of the enterprise.

[0008] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented in accordance with the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 A data storage optimization method flow diagram for business integration provided by an embodiment of the present application;

[0010] Figure 2 A data storage optimization platform structure diagram for business integration provided by an embodiment of the present application.

[0011] Explanation of reference numerals: department architecture extraction module 10, business data calling module 20, business interaction analysis module 30, historical business data segmentation module 40, topology network function enhancement module 50, intelligent data routing module 60. DETAILED DESCRIPTION

[0012] The embodiments of the present application provide a data storage optimization method and platform for business integration, which solves the technical problem that the data management and storage of different business departments in the prior art are usually static, isolated, lack of intelligent and dynamic optimization capability, and the efficiency of the overall business process is low.

[0013] The basic principles of the present application are introduced, the various non-limiting embodiments of the present application will be described in detail below in conjunction with the accompanying drawings. It should be understood that the specific examples described herein are merely intended to explain the present application and not to limit the present application.

[0014] Embodiment one, as shown, the present application provides a data storage optimization method for business integration, the method comprises: Figure 1

[0015] The department architecture of the target enterprise is extracted to obtain a plurality of business function departments.

[0016] The department architecture is the structure of the internal organization departments and their relationships of the enterprise, including the name, function, level, etc. of the department. The organizational structure information of the enterprise is obtained through the internal system of the target enterprise, including the organization architecture diagram, department responsibility description, etc. According to the organization architecture diagram and the department responsibility description, each department in the enterprise and the relationship between them are identified, and it is determined which departments are directly involved in the core business activities of the enterprise. A plurality of business function departments are obtained. The business function department is a department divided by the enterprise according to its main business function, specifically a department directly involved in the business process and core operation, such as the R&D department, the technical department, etc.

[0017] The business data of the plurality of business function departments is called to obtain a plurality of historical business data.

[0018] The extraction time range of the historical data is determined, such as the data of the past year, quarter or month. The extraction time range is determined according to the specific needs. The data source of each business function department is obtained, such as the business database within the department, etc. The data is extracted from the business database of each business function department according to the extraction time range to obtain a plurality of historical business data, which provides an accurate data basis for subsequent data analysis and business collaboration.

[0019] Based on the plurality of historical business data, the business interaction analysis of the plurality of business function departments is performed to construct and generate a data integration topology network, wherein the data integration topology network comprises a plurality of business data upload nodes mapped to the plurality of business function departments and K intermediates, and the K intermediates are used for pre-storing business integration data.

[0020] According to the plurality of historical business data, the flow path of the data between different business function departments is analyzed, including identifying which data is generated by which department, and how they are transmitted and shared between departments. By analyzing the frequency, direction and mode of data flow, the business interaction relationship between different departments is identified, for example, the sales department may regularly share sales data with the finance department and share inventory data with the warehousing department.

[0021] ​According to the results of business interaction analysis, a business data upload node is created for each business function department, which represents the position of each department in the data integration network and is the main interface for data inflow and outflow. According to the historical data flow path and the interaction between departments, the business data upload nodes are connected to obtain a business interaction topology network. In the business interaction topology network, K middleware is deployed, where K is a positive integer. The role of middleware is to serve as a temporary storage point for data pre-storage, helping to coordinate data flow between different business function departments.

[0022] According to the data flow direction of the plurality of historical business data, the plurality of historical business data is divided into a plurality of business data subsets, wherein each business data subset in the plurality of business data subsets includes a local data subset and a plurality of cross-department data subsets.

[0023] Through analysis of the plurality of historical business data, the flow direction of data between business function departments is identified, including analyzing the starting point (generation department), flow path (cross-department data transmission), and end point (receiving department) of the data, and determining the data flow direction.

[0024] According to the data flow direction, the plurality of historical business data is divided into a plurality of business data subsets, each data subset containing data with similar attributes under certain conditions. Each business data subset is composed of a local data subset and a plurality of cross-department data subsets. The local data subset is data generated within a certain business function department and mainly used by the department. These data do not need or rarely need to be shared with other departments. The cross-department data subset is data generated in one department but needs to be transmitted and shared between multiple departments. These data involve inter-department cooperation and business processes.

[0025] Based on the plurality of business data subsets, a plurality of business collaboration judgment models are constructed, and the plurality of business collaboration judgment models are mapped and synchronized to the plurality of business data upload nodes, completing the function enhancement of the data integration topology network.

[0026] Based on the plurality of business data subsets, a plurality of business collaboration judgment models are constructed, which are responsible for automatically classifying and organizing data to ensure that data is stored according to predetermined rules and business requirements. The specific construction process of the model is detailed in the subsequent steps. The constructed plurality of business collaboration judgment models are mapped to the corresponding business data upload nodes, each node is responsible for processing a specific business data subset, and performs data processing tasks according to the judgment results of the model. The model can be synchronized in real time with the business data upload node. By deploying business collaboration judgment models in the data integration topology network, new data processing functions are introduced, improving the efficiency and intelligence of the network.

[0027] In the data integration topology network, the multiple business data upload nodes predict the data flow direction of the business data dynamically uploaded by the multiple business function departments through the multiple business collaboration judgment models, and intelligently route the data based on the prediction results, completing the data storage update of the K intermediates.

[0028] In the daily operation process, multiple business function departments will generate new real-time business data. These real-time business data are uploaded to the corresponding multiple business data upload nodes by the respective business function departments. The business data upload nodes use the constructed business collaboration judgment model to predict the flow direction of the dynamically uploaded real-time business data. The model predicts the data flow path according to the learned features and patterns, and the output prediction results include collaborative data and non-collaborative data.

[0029] If it is collaborative data, the real-time flow department is obtained, and the real-time intermediate is located in the K intermediates according to the real-time flow department, and the real-time business data is routed to the real-time intermediate. If it is non-collaborative data, the real-time business data is formatted to obtain first formatted data, and the first formatted data is stored locally in the first business data upload node.

[0030] Through the above steps, the data can automatically flow to the correct processing node or storage location according to its type and the needs of the business department, realizing efficient data management and transmission, and thus improving the intelligent level of the overall business operation of the enterprise.

[0031] Further, based on the multiple historical business data, the business interaction analysis of the multiple business function departments is performed, and a data integration topology network is constructed. The method further comprises:

[0032] By performing data flow analysis on the multiple historical business data, a plurality of flow department sets are obtained. The multiple business data upload nodes are configured according to the multiple business function departments. The communication connection of the multiple business data upload nodes is performed according to the multiple flow department sets, and a business interaction topology network is obtained, wherein the business interaction topology network comprises K communication connection paths. In the business interaction topology network, the K communication connection paths are deployed with intermediates, and the data integration topology network is obtained.

[0033] The multiple historical business data are analyzed, the flow path of the data between the business function departments is tracked, and the complete flow path of the data is determined by identifying the generation department, the transmission department and the final receiving department of the data.

[0034] The plurality of historical business data is analyzed, the flow path of data between various business function departments is tracked, the complete flow path of data is determined by identifying the generation department, transmission department and final receiving department of the data, the dependency relationship between departments is identified according to the data flow path, for example, sales data first flows from the sales department to the finance department, and then to the logistics department, and this dependency relationship reflects the business contact strength between departments. Based on the data flow analysis result, a plurality of flow department sets are constructed, and each flow department set includes a plurality of business function departments participating in data flow, which jointly act in a specific business process.

[0035] The node configuration is performed, specifically, first, the number of business data upload nodes to be configured and their positions in the network are determined according to the number of business function departments and the data flow analysis result, then the main business requirements and data transmission requirements of each business function department are determined, for example, the sales department needs to frequently upload sales records, and the finance department needs to process a large amount of financial statement data, and the computing resources, storage space and network bandwidth required by the node are configured to ensure that it can meet the business requirements, and finally a plurality of configured business data upload nodes are obtained.

[0036] According to the plurality of flow department sets, the connection relationship between the departments is determined, and the communication connection between each business data upload node and other nodes is performed, which includes establishing the connection between the nodes at the physical and logical levels to ensure smooth flow of data between the nodes. After all the connection paths are established, the repeated paths are integrated to obtain a business interaction topology network, the business interaction topology network includes K communication connection paths, K is the number of communication connection paths in the business interaction topology network, and K is a positive integer.

[0037] The middleware is a component deployed in the network for optimizing data transmission, processing and storage, which can relieve node load, improve data transmission speed and enhance the overall function of the network. The K communication connection paths are deployed with middleware to optimize the connection paths between nodes and prevent single path overload. After the middleware deployment is completed, a complete data integration topology network is formed, which can not only process the transmission of business data, but also process and store data through the middleware.

[0038] Further, by performing data flow analysis on the plurality of historical business data, a plurality of flow department sets are obtained, and the method further comprises:

[0039] The plurality of historical business data is subjected to data flow path tracking to obtain a plurality of out-degree business data; the plurality of business function departments are taken as analysis starting points, the plurality of out-degree business data is traversed to obtain a plurality of discrete department calling sequences; and the plurality of discrete department calling sequences are subjected to business function department aggregation to obtain the plurality of flow-through department sets.

[0040] The plurality of historical business data is analyzed, focusing on external sharing of department data, tracking how data generated by each department flows to other departments, out-degree business data refers to data generated by one department and transmitted to other departments, and a plurality of out-degree business data sharing across departments is screened out, which is the basis for business integration between different departments in an enterprise and reflects the data dependency and collaboration relationship between departments.

[0041] Each business function department is taken as an analysis starting point, the analysis starting point department refers to the source department of data flow, i.e., the department that initially generates and shares data, each analysis starting point department is traversed one by one to track the flow of out-degree business data in other departments, and according to the order of data flow, a discrete department calling sequence corresponding to each out-degree data is generated, which shows the flow trajectory of data within an enterprise and reveals the data dependency relationship between departments.

[0042] The business function departments in the plurality of discrete department calling sequences are aggregated to form a plurality of flow-through department sets, the flow-through department set refers to a collection formed by multiple departments through data flow and interaction in a certain type of business process.

[0043] Further, based on the plurality of business data subsets, a plurality of business collaboration judgment models are constructed, and the method further comprises:

[0044] The first out-degree business data and the first flow-through department set are obtained by mapping data calling of the first business function department on the plurality of out-degree business data and the plurality of flow-through department sets, wherein the first business function department is any one of the plurality of business function departments; the first out-degree business data is split according to the first flow-through department set to obtain K cross-department data subsets, wherein the first flow-through department set includes K flow-through departments; the first historical business data mapped to the first business function department is extracted from the plurality of historical business data, and the first local data subset is obtained by excluding the first out-degree business data from the first historical business data; the first local data subset and the K cross-department data subsets are taken as a first group of business data subsets to construct a first business collaboration judgment model; and in the same way, the plurality of business collaboration judgment models are constructed based on the plurality of business data subsets.

[0045] Firstly, an analysis object is randomly extracted from a plurality of business function departments as a first business function department, and a mapping data call of the first business function department is performed in a plurality of out-degree business data and a plurality of flow-through department set to obtain first out-degree business data and a first flow-through department set directly related to the first business function department.

[0046] The first flow-through department set is analyzed to identify specific K flow-through departments contained in the set, which are key nodes that need to be transmitted across departments in the business integration process. According to the K flow-through departments, the first out-degree business data is split into K cross-department data subsets, each of which has a mapping relationship with the corresponding flow-through department.

[0047] From a plurality of historical business data, first historical business data corresponding to the first business function department is extracted, and the first out-degree business data is excluded from the first historical business data to distinguish which data is shared externally and which data is used only within the department. After excluding the out-degree business data, the remaining data is the first local data subset of the department, which is used only within the first business function department and does not involve cross-department business interaction.

[0048] Based on a multi-layer perception machine, a standard collaboration judgment model is constructed, the input layer of the model receives the features of the K cross-department data subsets, the hidden layer processes these features, and the output layer gives the judgment result. After model training, a business collaboration judgment layer is obtained. Based on the first local data subset, a data formatting layer is constructed. Cascading the business collaboration judgment layer and the data formatting layer, a first business collaboration judgment model is obtained, which simulates the data interaction habits of multiple business departments in business collaboration to judge whether the subsequent business upload data is applied to business department business collaboration.

[0049] By analogy, a plurality of business collaboration judgment models are constructed by traversing a plurality of business data subsets.

[0050] Further, the first local data subset and the K cross-department data subsets are used as a first group of business data subsets to construct a first business collaboration judgment model, and the method further comprises:

[0051] The K cross-department data subsets are subjected to feature vector extraction to obtain K business collaboration feature sets; a standard collaboration judgment model is built based on a multi-layer perception mechanism, and the K business collaboration feature sets and the K cross-department data subsets are used for model differentiation of the standard collaboration judgment model to obtain K business collaboration judgment branches; the K business collaboration judgment branches are connected in parallel to obtain a business collaboration judgment layer; the first local data subset is divided based on data reception time to obtain multiple update data subsets; the multiple update data subsets are subjected to format feature analysis to obtain a standard data format; a data formatting layer is built based on the standard data format; and the business collaboration judgment layer and the data formatting layer are cascaded to complete the construction of the first business collaboration judgment model.

[0052] First, explicit information (such as date, time, amount, etc.) and implicit information (such as transaction mode, behavior trend, etc.) are extracted from each cross-department data subset, which reflects the specific performance and potential rules of data in cross-department business. Through the fusion of explicit information and implicit information, feature vectors are generated, and K business collaboration feature sets are integrated, which contain the core features of data and provide a basis for subsequent model training and judgment.

[0053] Based on the extracted explicit and implicit information features, a standard collaboration judgment model is built using a multi-layer perception mechanism (MLP), which can predict the flow and processing of business data based on learned data features. According to specific business needs and the characteristics of cross-department data subsets, the standard model is differentiated to generate K business collaboration judgment branches that specifically handle specific types of data. Each branch focuses on processing different categories of cross-department data to ensure that the model can make targeted judgments and predictions.

[0054] The K business collaboration judgment branches are connected in parallel to form a business collaboration judgment layer, which can handle multiple data subsets in parallel and provide comprehensive business collaboration support by integrating the judgment results of each branch.

[0055] The time stamp of the data reception time is extracted from the first local data subset, which is the specific time point of data reception. According to business needs, define the division rules of data reception time, such as dividing by day, week, etc. According to the time window, the first local data subset is divided into multiple update data subsets, each containing data received within a specific time period.

[0056] The format feature analysis is performed on each updated data subset to identify the key attributes of the data format, including data type (such as text, number, date), name of data field, length of field, arrangement order of data, etc. Based on the analysis results, a unified standard data format is developed, which can cover common features in all updated data subsets, such as unifying the date format to YYYY-MM-DD, to ensure the operability and universality of the format.

[0057] A data formatting layer is designed to encode the rules of the standard data format into the formatting layer, and the data formatting layer is divided into several modules, each module handling different aspects of data formatting, such as different modules responsible for data type conversion, field standardization, date format unification, etc. The data formatting layer can be strengthened to ensure that minor differences in data are captured during the formatting process, further enhancing the processing capacity of the model.

[0058] The hierarchical relationship between the business collaboration judgment layer and the data formatting layer is determined. The data is first judged by the business collaboration judgment layer. If the business collaboration judgment result is a non-empty set, the real-time flow department is output. Otherwise, if the business collaboration judgment result is an empty set, the data formatting layer is activated for formatting processing, and then local storage is performed. According to this logic, the cascade of the business collaboration judgment layer and the data formatting layer is performed to ensure that the data is processed in the correct order by each layer.

[0059] The final first business collaboration judgment model realizes the in-depth understanding and accurate processing of enterprise business data through the iterative correlation learning mechanism. The core of the model is to extract and correlate explicit and implicit information, simulate data collaboration and flow between different business departments, and use multi-layer perceptron to predict and decide the flow direction based on data characteristics. Through model differentiation, multiple judgment branches are generated for different business needs and data types to achieve more detailed processing.

[0060] Further, feature vectors are extracted from the K cross-department data subsets to obtain K business collaboration feature sets. The method further comprises:

[0061] Explicit information is extracted from the K cross-department data subsets to obtain K explicit information sets. Implicit information is mined from the K cross-department data subsets to obtain K implicit information sets. The K explicit information sets and K implicit information sets both use the K flow department identifiers. Feature vectors are extracted based on the K explicit information sets and K implicit information sets to obtain the K business collaboration feature sets.

[0062] For each cross-department data subset, first extract the explicit information in the cross-department data, such as timestamp, transaction amount, product ID, etc., and encode it as explicit information, learn and extract these features through the encoder, and form the initial explicit feature vector. Use information association to identify and capture the detailed connections between explicit information, which reflect the relevance of data in different business processes, for example, the association between sales data and inventory data in a certain time period.

[0063] Through BP neural network, implicit information mining is performed on the cross-department data subset. Implicit information is derived through deep analysis of data, such as transaction patterns and user behavior trends. These information is not easy to observe directly, but is crucial for business collaboration judgment. Through iterative learning, information association is used to capture the relevance between implicit information and explicit information. This iterative learning process can better understand the relevance of data in different business scenarios.

[0064] Among them, the K explicit information sets and the K implicit information sets use corresponding K flow-through department identifiers. The feature vectors of the K explicit information sets and the K implicit information sets are integrated to obtain the K business collaboration feature sets.

[0065] Further, the method further comprises:

[0066] When the first business function department generates first real-time business data, the first business function department uploads the first real-time business data to the first business data upload node in the data integration topology network. In the first business data upload node, the K business collaboration judgment branches of the business collaboration judgment layer in the first business collaboration judgment model are used to perform synchronous data flow prediction. If the business collaboration judgment result is a non-empty set, output the real-time flow-through department. The first business data upload node locates a real-time middleware among the K middleware according to the real-time flow-through department, and routes the first real-time business data to the real-time middleware.

[0067] When the first business function department performs its regular business operation, it generates first real-time business data. The specific data content depends on the nature of the department's business. The first business function department transmits the generated first real-time business data to the corresponding first business data upload node through the network.

[0068] When the first service data uploading node receives the first real-time service data, the service cooperation judgment layer in the first service cooperation judgment model is activated, and K service cooperation judgment branches in the service cooperation judgment layer are called, the K service cooperation judgment branches synchronously judge and predict the data, each branch model predicts the flow direction of the data in the network according to the characteristics of its design and training, for example, some data may be determined to be transmitted to the finance department, and other data may be transmitted to the customer service department, and the final service cooperation judgment result is obtained by comprehensively considering the prediction results of the K branches.

[0069] If the service cooperation judgment result is a non-empty set, it means that the data needs to be transmitted to one or more flow departments for processing, the non-empty set judgment result includes a group of flow departments, which are the next receivers of the service data, and the real-time flow departments receiving the data are identified according to the output of the model.

[0070] According to the real-time flow department, the middleware associated with the department is located in the network, each middleware is associated with a flow department and is responsible for temporarily storing, processing or forwarding service data, and the first real-time service data is transmitted from the uploading node to the located real-time middleware, and the data is temporarily stored, format converted, data aggregated or further processed.

[0071] Further, the method further comprises:

[0072] If the service cooperation judgment result is a non-empty set, the data formatting layer of the first service cooperation judgment model is activated, the first real-time service data is processed by the data formatting layer, first formatted data is obtained, and the first formatted data is stored locally in the first service data uploading node.

[0073] If the service cooperation judgment result is a non-empty set, it means that the current data does not need to be transmitted to other flow departments for processing, in this case, the data formatting layer is activated to standardize the data, and the data is stored locally instead of being continuously transmitted.

[0074] According to the type and content of the first real-time service data, the applicable formatting rules are selected, which define how to process each part of the data, including date format, numerical range, field sorting, etc., and the data formatting layer processes each piece of first real-time service data according to the formatting rules to ensure the uniformity and standardization of the data format, and generates first formatted data after processing operation. The first formatted data is stored locally in the first service data uploading node.

[0075] In summary, the data storage optimization method for service integration provided by the embodiments of the present application has the following technical effects:

[0076] By extracting the department architecture of the target enterprise and calling the historical business data of each department, the business flow within the enterprise can be comprehensively understood, and the interaction relationship and data dependency between departments can be effectively identified, laying a foundation for subsequent data integration; based on the historical business data, a data integration topology network is constructed, which effectively reflects the interaction relationship between each business function department within the enterprise, and through mapping multiple business data upload nodes and K intermediates in the network, the pre-storage and management of business data are realized, such topology structure not only optimizes the distribution of data storage, but also improves the speed and efficiency of data access, ensuring the efficient flow of data between departments; by analyzing the data flow of historical business data, the data is divided into multiple sets of business data subsets, and on this basis, multiple business collaboration judgment models are constructed, which can deeply analyze the explicit and implicit information of business data and identify the relevance between data, such modeling design enables the system to intelligently judge whether the business data needs to be uploaded or flow between different departments, thereby enhancing the collaboration ability and decision support between departments; at the data upload node, the business collaboration judgment model is used to predict the flow direction of real-time uploaded data, and intelligent data routing is performed based on the prediction results, ensuring that the data can be timely and accurately transmitted to the appropriate intermediate for processing or storage, thereby optimizing the resource utilization of the enterprise.

[0077] In the second embodiment, based on the same inventive concept as the data storage optimization method for business integration in the preceding embodiments, as shown in the following Figure 2 The embodiments of the present application provide a data storage optimization platform for business integration, which comprises:

[0078] The department architecture extraction module 10 is used for department architecture extraction of a target enterprise, and a plurality of business function departments are obtained. The business data calling module 20 is used for business data calling of the plurality of business function departments, and a plurality of historical business data are obtained. The business interaction analysis module 30 is used for business interaction analysis of the plurality of business function departments based on the plurality of historical business data, and a data integration topology network is constructed and generated. The data integration topology network includes a plurality of business data upload nodes and K intermediates mapped to the plurality of business function departments. The K intermediates are used for pre-storing business integration data. The historical business data segmentation module 40 is used for segmenting the plurality of historical business data into a plurality of business data subsets according to data flow directions of the plurality of historical business data. Each business data subset in the plurality of business data subsets includes a local data subset and a plurality of cross-department data subsets. The topology network function enhancement module 50 is used for constructing a plurality of business cooperation judgment models based on the plurality of business data subsets, and mapping and synchronizing the plurality of business cooperation judgment models to the plurality of business data upload nodes, so as to complete function enhancement of the data integration topology network. The intelligent data routing module 60 is used for, in the data integration topology network, the plurality of business data upload nodes performing data flow direction prediction on business data dynamically uploaded by the plurality of business function departments through the plurality of business cooperation judgment models, and performing intelligent data routing based on a prediction result, so as to complete data storage update of the K intermediates.

[0079] Further, the platform further includes a data integration topology network construction module to perform the following operation steps:

[0080] By performing data flow analysis on the plurality of historical business data, a plurality of flow transfer department sets are obtained. The plurality of business data upload nodes are configured according to the plurality of business function departments. The plurality of business data upload nodes are communicatively connected according to the plurality of flow transfer department sets, and a business interaction topology network is obtained. The business interaction topology network includes K communication connection paths. In the business interaction topology network, the K communication connection paths are deployed with intermediates, and the data integration topology network is obtained.

[0081] Further, the platform further includes a plurality of flow transfer department set acquisition modules to perform the following operation steps:

[0082] The plurality of historical business data are subjected to data flow path tracking to obtain a plurality of out-degree business data; the plurality of business function departments are taken as analysis starting points, and the plurality of out-degree business data are traversed to obtain a plurality of discrete department calling sequences; the plurality of discrete department calling sequences are subjected to business function department aggregation to obtain the plurality of flow-through department sets.

[0083] Further, the platform further comprises a business cooperation judgment model construction module to perform the following operation steps:

[0084] The plurality of out-degree business data and the plurality of flow-through department sets are subjected to mapping data calling based on a first business function department to obtain first out-degree business data and a first flow-through department set, wherein the first business function department is any one of the plurality of business function departments; the first out-degree business data is split according to the first flow-through department set to obtain K cross-department data subsets, wherein the first flow-through department set comprises K flow-through departments; first historical business data mapped to the first business function department is extracted from the plurality of historical business data, and a first local data subset is obtained by excluding the first out-degree business data from the first historical business data; the first local data subset and the K cross-department data subsets are taken as a first group of business data subsets to construct a first business cooperation judgment model; and by analogy, the plurality of business cooperation judgment models are constructed based on the plurality of groups of business data subsets.

[0085] Further, the platform further comprises a first business cooperation judgment model construction module to perform the following operation steps:

[0086] Feature vectors are extracted from the K cross-department data subsets to obtain K business cooperation feature sets; a standard cooperation judgment model is constructed based on a multilayer perception mechanism, the K business cooperation feature sets and the K cross-department data subsets are subjected to model differentiation of the standard cooperation judgment model to obtain K business cooperation judgment branches; the K business cooperation judgment branches are connected in parallel to obtain a business cooperation judgment layer; the first local data subset is divided based on data receiving time to obtain a plurality of updated data subsets; the plurality of updated data subsets are subjected to format feature analysis to obtain a standard data format; a data formatting layer is constructed based on the standard data format; and the business cooperation judgment layer and the data formatting layer are cascaded to complete construction of the first business cooperation judgment model.

[0087] Further, the platform further comprises a K business cooperation feature set acquisition module to perform the following operation steps:

[0088] The K cross-department data subsets are subjected to explicit information extraction to obtain K explicit information sets; the K cross-department data subsets are subjected to implicit information mining to obtain K implicit information sets, wherein the K explicit information sets and the K implicit information sets are both subjected to the K flow transfer department identifiers; feature vector extraction is performed based on the K explicit information sets and the K implicit information sets to obtain the K business cooperation feature sets.

[0089] Further, the platform further comprises a first real-time business data routing module to perform the following operation steps:

[0090] When the first business function department generates first real-time business data, the first business function department uploads the first real-time business data to a first business data uploading node in the data integration topology network; at the first business data uploading node, data flow direction prediction is synchronously performed through K business cooperation judgment branches of the business cooperation judgment layer in the first business cooperation judgment model; if the business cooperation judgment result is a non-empty set, a real-time flow transfer department is output; the first business data uploading node positions a real-time middleware according to the real-time flow transfer department among the K middlewares and routes the first real-time business data to the real-time middleware.

[0091] Further, the platform further comprises a local storage module to perform the following operation steps:

[0092] If the business cooperation judgment result is an empty set, the data formatting layer of the first business cooperation judgment model is activated; the first real-time business data is subjected to formatting processing through the data formatting layer to obtain first formatted data; the first formatted data is subjected to local storage at the first business data uploading node.

[0093] Through the foregoing detailed description of the data storage optimization method for business integration, those skilled in the art can clearly understand the data storage optimization platform for business integration in the embodiment. Since the platform corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the method part description.

[0094] The foregoing description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data storage optimization method for business integration, characterized in that, The method includes: Extract the departmental structure of the target company to obtain multiple business function departments; The business data of the multiple business function departments is called to obtain multiple historical business data. Based on the aforementioned historical business data, business interaction analysis is performed on the aforementioned business functional departments to construct and generate a data integration topology network. The data integration topology network includes multiple business data upload nodes mapped to the aforementioned business functional departments and K middleware. The K middleware is used for pre-storing business integration data. Based on the data flow of the multiple historical business data, the multiple historical business data are divided into multiple sets of business data subsets, wherein each set of business data subsets includes a local data subset and multiple cross-department data subsets; Multiple business collaboration judgment models are constructed based on the multiple sets of business data subsets, and the multiple business collaboration judgment models are mapped and synchronized to the multiple business data upload nodes to complete the functional enhancement of the data integration topology network. In the data integration topology network, the multiple business data upload nodes predict the data flow direction of the business data dynamically uploaded by the multiple business function departments through the multiple business collaboration judgment models, and perform intelligent data routing based on the prediction results to complete the data storage and update of the K middleware.

2. The data storage optimization method for business integration as described in claim 1, characterized in that, Based on the aforementioned historical business data, the method performs business interaction analysis on the multiple business functional departments to construct and generate a data integration topology network. The method further includes: By performing data flow analysis on the aforementioned historical business data, multiple sets of flow departments are obtained; Configure the multiple business data upload nodes according to the multiple business function departments; Based on the multiple sets of circulation departments, communication connections are established between the multiple business data upload nodes to obtain a business interaction topology network, wherein the business interaction topology network includes K communication connection paths; In the business interaction topology network, middleware is deployed on the K communication connection paths to obtain the data integration topology network.

3. The data storage optimization method for business integration as described in claim 2, characterized in that, By performing data flow analysis on the aforementioned historical business data, multiple sets of flow departments are obtained. The method further includes: Data flow path tracing is performed on the multiple historical business data to obtain multiple out-degree business data; Using the aforementioned multiple business function departments as the starting point for analysis, the multiple out-degree business data are traversed to obtain multiple discrete department call sequences; The business function department is aggregated for the multiple discrete department call sequences to obtain the multiple flow department sets.

4. The data storage optimization method for business integration as described in claim 3, characterized in that, The method further includes constructing multiple business collaboration judgment models based on the aforementioned subsets of business data, and further includes: Based on the mapping data call of the multiple outgoing business data and multiple circulation department sets by the first business function department, the first outgoing business data and the first circulation department set are obtained, wherein the first business function department is any one of the multiple business function departments; The first outgoing business data is split according to the first set of circulation departments to obtain K cross-departmental data subsets, wherein the first set of circulation departments includes K circulation departments; First historical business data mapped to the first business function department is extracted from the plurality of historical business data, and a first local data subset is obtained by removing the first outgoing business data from the first historical business data; The first local data subset and K cross-departmental data subsets are used as the first group of business data subsets to construct the first business collaboration judgment model; Similarly, multiple business collaboration judgment models are constructed based on the multiple subsets of business data.

5. The data storage optimization method for business integration as described in claim 4, characterized in that, The method further includes constructing a first business collaboration judgment model using the first local data subset and K cross-departmental data subsets as the first group of business data subsets. Feature vectors are extracted from the K cross-departmental data subsets to obtain K business collaboration feature sets; A standard collaboration judgment model is constructed based on a multilayer perceptron. The model is differentiated using the K business collaboration feature sets and K cross-departmental data subsets to obtain K business collaboration judgment branches. The K business collaboration decision branches are connected in parallel to obtain the business collaboration decision layer; The first local data subset is segmented based on the data reception time to obtain multiple updated data subsets; Format feature analysis is performed on the multiple updated data subsets to obtain a standard data format; A data formatting layer is constructed based on the aforementioned standard data format; The business collaboration judgment layer and the data formatting layer are cascaded to complete the construction of the first business collaboration judgment model.

6. The data storage optimization method for business integration as described in claim 5, characterized in that, The method further includes extracting feature vectors from the K cross-departmental data subsets to obtain K business collaboration feature sets. Explicit information extraction is performed on the K cross-departmental data subsets to obtain K explicit information sets; Implicit information mining is performed on the K cross-departmental data subsets to obtain K implicit information sets, wherein both the K explicit information sets and the K implicit information sets adopt K circulation department identifiers; Feature vectors are extracted based on the K explicit information sets and K implicit information sets to obtain the K business collaboration feature sets.

7. The data storage optimization method for business integration as described in claim 5, characterized in that, The method further includes: When the first business function department generates the first real-time business data, the first business function department uploads the first real-time business data to the first business data upload node in the data integration topology network. At the first business data upload node, data flow prediction is performed synchronously through the K business collaboration judgment branches of the business collaboration judgment layer in the first business collaboration judgment model. If the business collaboration judgment result is a non-empty set, then output the real-time transfer department; The first business data upload node locates the real-time middleware among the K middlewares according to the real-time flow department, and routes the first real-time business data to the real-time middleware.

8. The data storage optimization method for business integration as described in claim 7, characterized in that, The method further includes: If the business collaboration judgment result is an empty set, then the data formatting layer of the first business collaboration judgment model is activated; The first real-time business data is formatted through the data formatting layer to obtain first formatted data; The first formatted data is stored locally on the first business data upload node.

9. A data storage optimization platform for business integration, characterized in that, The platform is used to implement the data storage optimization method for business integration as described in any one of claims 1-8, and comprises: The departmental structure extraction module is used to extract the departmental structure of the target enterprise and obtain multiple business function departments. A business data retrieval module is used to retrieve business data from the multiple business function departments and obtain multiple historical business data. A business interaction analysis module is used to perform business interaction analysis on the multiple business functional departments based on the multiple historical business data, and to construct and generate a data integration topology network. The data integration topology network includes multiple business data upload nodes mapped to the multiple business functional departments and K middleware. The K middleware is used for pre-storing business integration data. A historical business data segmentation module is used to segment the multiple historical business data into multiple business data subsets according to the data flow direction of the multiple historical business data. Each of the multiple business data subsets includes a local data subset and multiple cross-department data subsets. A topology network function enhancement module is used to construct multiple business collaboration judgment models based on the multiple sets of business data subsets, and to map and synchronize the multiple business collaboration judgment models to the multiple business data upload nodes, thereby completing the function enhancement of the data integration topology network. The intelligent data routing module is used in the data integration topology network to predict the data flow direction of the business data dynamically uploaded by the multiple business function departments through the multiple business collaboration judgment models, and to perform intelligent data routing based on the prediction results, thereby completing the data storage and update of the K middleware.

Images