Big data classification utilization method and system for business management

Abstract

The invention provides a big data classification utilization method and system for business management. The method comprises the following steps: receiving real-time data streams from an enterprise internal information system and an external public data source; automatically identifying key information in the real-time data flow according to a preset service rule, and mapping the key information to a corresponding activity category; integrating the key information into a unified database, and generating a corresponding metadata tag for each category of key information; carrying out rapid classification processing on the key information with the metadata labels by utilizing a distributed computing framework to form a structured data set; providing an access interface of the structured data set for an authorized user through an authority management system; and in response to the data analysis request, dynamically adjusting the presentation form of the structured data set. According to the technical scheme provided by the invention, the method not only improves the operation efficiency of the coal feeder, but also enhances the overall performance of the system, and provides a more intelligent and efficient solution for industrial production.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of big data classification and utilization for business management, and in particular, to a method and system for big data classification and utilization for business management. Background Art

[0002] In modern industrial production, especially in industries such as coal and electricity, coal feeders, as important material conveying equipment, their operating efficiency directly affects the efficiency and energy consumption of the entire production line. With the development of automation and intelligent technologies, remote monitoring and control have become one of the key means to improve equipment management efficiency.

[0003] Currently, most coal feeder control systems still rely on on-site manual adjustment or simple timing control. Although some advanced systems have begun to attempt remote control, the following problems generally exist: existing remote control systems often lack sufficient security measures and are vulnerable to network attacks or data tampering, resulting in unreliable transmission of control instructions; when existing remote control systems receive and execute frequency conversion instructions, the execution effect is often poor due to signal delay or instability, and the operating frequency of the coal feeder cannot be accurately adjusted; after adjusting the operating frequency, there is a lack of effective monitoring means to feedback the working state information of the coal feeder to the remote control center in real time, making it difficult for the remote control center to adjust the control strategy in a timely manner according to the actual situation.

[0004] The defects of the existing solutions are as follows: there are security risks in the communication connection between the remote control center and the coal feeder in the existing solutions, which are vulnerable to attacks, affecting the system stability and data integrity; the existing solutions cannot ensure the accurate execution of frequency conversion instructions, resulting in inaccurate adjustment of the operating frequency of the coal feeder and affecting production efficiency; the existing solutions lack an effective real-time monitoring and feedback mechanism and cannot adjust the control strategy in a timely manner to optimize the operating efficiency of the coal feeder. Summary of the Invention

[0005] The embodiments of the present application provide a method and system for big data classification and utilization for business management to solve the problem in the prior art that there is a lack of an effective real-time monitoring and feedback mechanism and the control strategy cannot be adjusted in a timely manner to optimize the operating efficiency of the coal feeder.

[0006] In a first aspect, the embodiments of the present application provide a method for big data classification and utilization for business management, including:

[0007] Receive real-time data streams from internal enterprise information systems and external public data sources; automatically identify key information in the real-time data streams according to preset business rules and map it to corresponding activity categories; integrate the key information into a unified database and generate corresponding metadata tags for each category of key information; use a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set; provide an access interface to the structured data set to authorized users through a permission management system, enabling users to implement customized data analysis requests based on the interface; in response to the data analysis request, dynamically adjust the presentation form of the structured data set to support the display of different-dimensional data views required in the decision-making process. Optionally, according to the method described in claim 1, the automatically identifying key information in the real-time data streams according to preset business rules and mapping it to corresponding activity categories includes: parsing the text content in the real-time data streams through natural language processing technology to identify first key information containing specific keywords or phrases; using predefined regular expressions or pattern matching algorithms to filter structured or semi-structured data items in the real-time data streams as second key information, the regular expressions or pattern matching algorithms being customized based on preset business rules to capture data items conforming to specific formats; analyzing the real-time data streams by applying a machine learning model based on the preset business rules to obtain third key information, the machine learning model being constructed through learning a large amount of historical data to identify potential and non-explicit key information; taking the first key information, the second key information, and the third key information as key information and mapping it to preset activity categories.

[0008] Optionally, the integrating the key information into a unified database and generating corresponding metadata tags for each category of key information includes:

[0009] Import the identified and classified key information into a unified database; create corresponding data tables or partitions for each category of key information in the database; use a metadata management system to assign unique identifiers to the key information and generate corresponding metadata tags according to the attributes of the key information, the attributes including at least source, type, creation time, and update time.

[0010] Optionally, the using a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set includes:

[0011] Using a distributed computing framework, initially group the key information according to metadata tags to obtain the result of the initial grouping; deploy a machine learning algorithm model in the distributed computing framework, and apply corresponding classification algorithms to the key information of different metadata tags. The classification algorithms are trained based on historical data and can effectively distinguish different types of data; perform secondary processing on the result of the initial grouping, merge adjacent or overlapping data groups, and ensure that the data within all groups belongs to the same category through a consistency check algorithm; use data compression technology to compress and store the classified data, reducing the storage cost while retaining the integrity and availability of the data to form a structured data set.

[0012] Optionally, in response to the data analysis request, dynamically adjust the presentation form of the structured data set to support the display of different-dimensional data views required in the decision-making process, including:

[0013] Receive a data analysis request submitted by a user through a front-end interface or an API interface. The data analysis request contains the data dimension information required by the user; according to the data dimension information in the data analysis request, extract relevant data subsets from the structured data set. The data subsets are filtered from the structured data set according to the data dimension information selected by the user; apply a data aggregation algorithm to aggregate the data subsets to generate a data summary result suitable for display; use a dynamic data visualization component to display the data summary result.

[0014] Optionally, applying the data aggregation algorithm to aggregate the data subsets to generate a data summary result suitable for display includes:

[0015] For a given data subset D = {d1, d2, …, d n}, each data point d i contains multiple attributes a ij , where j represents the jth attribute. The data aggregation algorithm is defined as a dynamic weight and spatio-temporal awareness data aggregation formula based on a multi-modal fusion deep reinforcement learning model:

[0016]

[0017] where P(D, A) represents the data aggregation result of the data subset D on the attribute set A; W j (d i , t) is the dynamic weight of the jth attribute of the data point d i at time t, and the dynamic weight is dynamically adjusted through a multi-modal fusion deep reinforcement learning model; f(a ij ) is a function that, according to the attribute a ijto calculate the corresponding aggregation value according to the type, α is the influence coefficient of the context factor, and the influence coefficient is dynamically adjusted by an adaptive algorithm according to the importance of the current context. G(Context(d i ),t) is a function that adjusts the final aggregation result according to the context information of the data point d i and its timestamp t. The function also learns the correlation between the context and the data point through a deep reinforcement learning model; γ is the influence coefficient of the spatial location factor, and this coefficient is also dynamically adjusted by an adaptive algorithm; H(Spatial(d i ),t) is a function that further adjusts the data aggregation result according to the spatial location information of the data point d i and its timestamp t. The function learns the correlation between the geographical location and the data through a deep reinforcement learning model; δ is the influence coefficient of the prediction result of the machine learning model; M(a ij ,t) is a prediction function based on time series analysis, which is used to predict the trend of a certain attribute a ij in the future time period. The function can be predicted using a deep learning model; θ is the influence coefficient of the reward mechanism; R(d i ,t) is a reward function that determines its contribution to the final aggregation result according to the importance of the data point d i at time t. The function is dynamically adjusted by a deep reinforcement learning model according to historical behavior; λ is the behavior influence coefficient; B(d i ,t) is a behavior function that adjusts the influence on the data aggregation result according to the behavior pattern associated with the data point d i .

[0018] Optionally, displaying the data summary result using a dynamic data visualization component includes:

[0019] Generating a visualization chart according to the data summary result P(D,A), and the visualization chart is used to reflect the change trend of the data over time, including line charts, bar charts, and heat maps;

[0020] The generation algorithm of the visualization chart of the dynamic data is defined as:

[0021]

[0022] Among them, V(P) represents the visualization chart of the data summary result P(D,A), is the visualization weight adjusted according to the user preference U; C i (C) is the context weight adjusted according to the context information C; S i (S) is the position weight adjusted according to the spatial location information S; T context (Tcontext ) is the time context weight adjusted according to the context time T context ; H i (H) is the historical behavior weight adjusted according to the historical behavior pattern H; U(U) is a function that selects the appropriate chart type according to the user preference U; T(T) is a function that automatically adjusts the time axis display of the chart according to the time dimension T.

[0023] In a second aspect, an embodiment of the present application provides a big data classification and utilization system for business management, including:

[0024] A receiving module, configured to receive real-time data streams from an enterprise internal information system and an external public data source; an identification module, configured to automatically identify key information in the real-time data streams according to preset business rules and map it to corresponding activity categories; an integration module, configured to integrate the key information into a unified database and generate corresponding metadata tags for each category of key information; a classification module, configured to perform fast classification processing on the key information with metadata tags by using a distributed computing framework to form a structured data set; a providing module, configured to provide an access interface to the structured data set to authorized users through a permission management system, so that users can implement customized data analysis requests based on the interface; an adjustment module, configured to dynamically adjust the presentation form of the structured data set in response to the data analysis request to support the display of different-dimensional data views required in the decision-making process.

[0025] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a big data classification and utilization method and system for business management as described in the first aspect.

[0026] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it implements a big data classification and utilization method and system for business management as described in the first aspect.

[0027] In the embodiments of the present application, real-time data streams are received from the enterprise internal information system and external public data sources; key information in the real-time data streams is automatically identified according to preset business rules and mapped to corresponding activity categories; the key information is integrated into a unified database, and corresponding metadata tags are generated for the key information of each category; the key information with metadata tags is quickly classified by using a distributed computing framework to form a structured data set; an access interface to the structured data set is provided to authorized users through a permission management system; in response to the data analysis request, the presentation form of the structured data set is dynamically adjusted. The technical solution provided by the present application not only improves the operation efficiency of the coal feeder, but also enhances the overall performance of the system, providing a more intelligent and efficient solution for industrial production.

[0028] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0030] Figure 1 It is a flowchart of a method for classifying and utilizing big data for business management provided by an embodiment of the present application;

[0031] Figure 2 It is a schematic structural diagram of a system for classifying and utilizing big data for business management provided by an embodiment of the present application;

[0032] Figure 3 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

[0034] In some processes described in the specification, claims, and the above-mentioned drawings of this application, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0035] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0036] Figure 1 The flowchart of a method for classifying and utilizing big data for business management provided by an embodiment of the present application is as Figure 1 shown, and the method includes:

[0037] 101. Receive real-time data streams from enterprise internal information systems and external public data sources;

[0038] This step refers to continuously collecting data from different data sources (including but not limited to enterprise internal ERP systems, CRM systems, financial systems, sales systems, etc., and external social media platforms, news websites, market research reports, etc.). These data usually exist in the form of streams, that is, the data does not arrive all at once, but new data points are continuously generated over time. The characteristics of real-time data streams are large data volume, high speed, and high diversity, so special technologies and tools are required for processing.

[0039] Suppose a retail company wishes to improve its inventory management and market response speed. Then the company can implement the following steps to receive real-time data streams: receive information on inventory levels, order status, production progress, etc.; collect customer behavior data such as purchase history, browsing records, customer feedback, etc.; track financial-related information such as transaction records, cash flow status, etc.; obtain point-of-sale (POS) data to understand which products sell best and when the sales peaks are, etc.; monitor discussions, evaluations, and sentiment towards the brand by consumers to understand market sentiment; track industry trends, latest moves of competitors, policy changes, etc.; subscribe to research reports of relevant industries to obtain market trends and forecast data; obtain local weather forecast data as weather conditions may affect the sales of certain products.

[0040] 102. Automatically identify the key information in the real-time data stream according to preset business rules and map it to the corresponding activity categories;

[0041] This step means that after receiving the real-time data stream, through preset business rules, automatically identify which data is important and classify this key information into specific activity categories. The purpose of doing this is to simplify the subsequent data processing process and ensure that the information processed is directly related to specific business objectives.

[0042] Suppose an e-commerce company wishes to improve its customer service experience and enhance the effectiveness of marketing activities. Then it can implement the following steps to identify the key information in the real-time data stream and map it to the corresponding activity categories: identify the keywords in customer inquiries such as "return", "refund", "delivery delay" and label them as the "customer service" category; scrape positive reviews and media reports related to the company's brand or products from social media and news websites and label them as the "brand promotion" category; identify product information with inventory below the warning threshold from the sales system and label it as the "inventory shortage" category; extract information on industry growth and emerging market opportunities from market research reports and label it as the "market intelligence" category; use NLP technology to parse text content such as customer service chat records, social media comments, etc. to identify the first key information containing specific keywords or phrases; use predefined regular expressions or pattern matching algorithms to screen structured or semi-structured data items such as order numbers, product IDs, etc. as the second key information; train a machine learning model based on a large amount of historical data to identify potential and non-explicit key information such as customers' purchase intentions or changes in market trends; map the identified first key information, second key information, and the third key information obtained through the machine learning model to preset activity categories such as customer service, marketing activities, inventory management, etc.

[0043] This application takes into account that in the daily operations of e-commerce companies, a large amount of real-time data streams (such as customer service chat records, social media comments, sales records, etc.) are continuously generated, which poses high requirements for the performance and accuracy of data processing systems. Although existing technical solutions can process some data, when faced with complex and variable data types, the following problems often occur: Due to the wide variety of data types, it is difficult for traditional manual or simple automation methods to comprehensively cover all important information; relying on a single technical means (such as keyword matching) may lead to inaccurate information extraction, especially for fuzzy or implicit information; in the face of massive data, without an efficient processing mechanism, it may lead to data processing lags and inability to respond to business needs in real time; how to accurately map the extracted key information to the corresponding business activities is a challenge, especially when the information itself does not directly point to a specific activity.

[0044] To solve the above technical problems, an alternative solution is proposed in an embodiment of the present invention. By combining natural language processing technology, pattern matching algorithms, and machine learning models, automatic identification of key information in real-time data streams is achieved and mapped to the corresponding activity categories, thereby improving the accuracy and efficiency of information extraction.

[0045] The alternative solution is as follows:

[0046] Optionally, the "automatically identifying the key information in the real-time data stream according to preset business rules and mapping it to the corresponding activity category" in step 102 includes: parsing the text content in the real-time data stream through natural language processing technology to identify the first key information containing specific keywords or phrases; using predefined regular expressions or pattern matching algorithms to filter structured or semi-structured data items in the real-time data stream as the second key information, and the regular expressions or pattern matching algorithms are customized based on preset business rules to capture data items that conform to specific formats; based on the preset business rules, applying a machine learning model to analyze the real-time data stream to obtain the third key information, and the machine learning model is constructed through learning a large amount of historical data to identify potential and non-explicit key information; taking the first key information, the second key information, and the third key information as key information and mapping them to the preset activity categories.

[0047] Suppose an e-commerce company hopes to improve the customer service experience and enhance the effectiveness of marketing activities. The specific implementation steps are as follows:

[0048] The company uses NLP technology to analyze customer service chat records, identify conversations containing words such as "return" and "refund", and label them as the "Customer Service" category; extract comments containing words such as "positive reviews" and "recommendations" from social media comments and label them as the "Brand Promotion" category; use regular expressions to filter order numbers and product IDs from sales records, and these data items will be labeled as "Sales Records"; apply a machine learning model trained with historical data to analyze customers' purchase behaviors, identify users with high purchase intentions, and label them as "Potential Customers"; map the identified key information (such as return requests, positive reviews, potential customers, etc.) to corresponding activity categories, such as "Customer Service", "Brand Promotion", and "Marketing Campaigns".

[0049] Combining multiple technical means to ensure comprehensive coverage and accurate extraction of key information; through comprehensive analysis, accurately map the information to the corresponding business activities to avoid incorrect information classification; utilize efficient natural language processing technology and machine learning models to achieve rapid response and processing of real-time data streams; provide a structured and clearly classified dataset to support more accurate business analysis and decision-making; the combined effects of these solve the problems of data omission, inaccurate information extraction, and low processing efficiency existing in the prior art, and significantly improve the data processing capabilities and business operation efficiencies of e-commerce companies in aspects such as customer service and marketing campaigns.

[0050] 103. Integrate the key information into a unified database and generate corresponding metadata tags for each category of key information;

[0051] This step refers to integrating the key information identified and classified in the previous step into a unified database and adding metadata tags to each piece of key information to facilitate subsequent data retrieval, management, and analysis. Metadata tags can help describe the characteristics of the data, such as the data source, type, creation time, etc., thus making the data more readable and manageable.

[0052] Suppose a manufacturing enterprise wants to improve its supply chain management efficiency and better track production and logistics. Then it can implement the following steps to integrate key information and generate metadata tags: from which system or sensor the data is collected; the nature of the data (such as production data, logistics data, inventory data, etc.); the specific timestamp when the data is generated; the timestamp when the data was last modified; the location where the data is collected or the geographical location information related to it; whether the data is associated with a specific entity (such as product model, supplier name, etc.); integrate the key information identified from the enterprise's internal information systems (such as ERP, CRM, SCM, etc.) and external public data sources (such as weather forecasts, market reports, etc.) into a unified database; for each piece of key information, assign metadata tags according to its characteristics.

[0053] This application considers that in the manufacturing industry, with the development of Internet of Things (IoT) technology, enterprises can collect a large amount of real-time data from various devices and systems. However, this data is often scattered across different systems and platforms, lacking unified management and organization, resulting in a serious data silo phenomenon, which is not conducive to the comprehensive utilization and analysis of data. In addition, due to the lack of an effective metadata management mechanism, enterprises often encounter difficulties in retrieving and using data, such as not knowing the data source, timeliness, and relevance.

[0054] To solve these problems, an alternative solution is proposed in the embodiments of the present invention. By integrating the identified and classified key information into a unified database and generating corresponding metadata tags for each category of key information, the readability and manageability of the data are improved, supporting more efficient data retrieval and analysis.

[0055] The alternative solution is as follows:

[0056] Optionally, the step of "integrating the key information into a unified database and generating corresponding metadata tags for each category of key information" in step 103 includes:

[0057] Importing the identified and classified key information into a unified database; creating corresponding data tables or partitions for each category of key information in the database; using a metadata management system to assign a unique identifier to the key information and generating corresponding metadata tags according to the attributes of the key information, where the attributes at least include source, type, creation time, and update time.

[0058] In the manufacturing industry, with the development of Internet of Things (IoT) technology, enterprises can collect a large amount of real-time data from various devices and systems. However, this data is often scattered across different systems and platforms, lacking unified management and organization, resulting in a serious data silo phenomenon, which is not conducive to the comprehensive utilization and analysis of data. In addition, due to the lack of an effective metadata management mechanism, enterprises often encounter difficulties in retrieving and using data, such as not knowing the data source, timeliness, and relevance.

[0059] To solve these problems, an alternative solution is proposed in the embodiments of the present invention. By integrating the identified and classified key information into a unified database and generating corresponding metadata tags for each category of key information, the readability and manageability of the data are improved, supporting more efficient data retrieval and analysis.

[0060] Suppose an enterprise needs to analyze the production situation of a batch of products. The specific steps are as follows: Query the production data records related to the product model "ModelXYZ" in the database; Conduct data analysis based on the query results, such as counting the production quantity and average production rate during this period;

[0061] Calculate the average production rate using statistical formulas:

[0062]

[0063] Suppose the total production quantity is 1000 pieces and the total production time is 8 hours. Then the average production rate is:

[0064]

[0065] By implementing this optional solution, manufacturing enterprises can achieve the following beneficial effects: Integrate scattered data into a unified database for easy centralized management and unified access; Through the generation of metadata tags, make the data more readable and manageable, support more efficient data retrieval and analysis; Through a unified data organization method, promote cross-departmental and cross-system data sharing and utilization, and improve the overall operation efficiency of the enterprise; Provide a structured and clearly classified dataset to support management decision-making based on accurate data; These effects work together to solve problems such as data islands and difficult retrieval in the existing technology, and significantly improve the data processing ability and business operation efficiency of manufacturing enterprises in supply chain management.

[0066] 104. Use a distributed computing framework to quickly classify and process key information with metadata tags to form a structured dataset;

[0067] The goal of this step is to process the key information that has been integrated and tagged with metadata tags before through a distributed computing framework (such as Apache Hadoop, Apache Spark, etc.). The distributed computing framework can distribute processing tasks among multiple computers, thus accelerating the processing speed of large-scale data. Through this process, the original unstructured data will be converted into a structured dataset for further data analysis and mining.

[0068] Suppose a financial services company wants to improve its risk management capabilities and better understand customer behavior by analyzing customer transaction records. Then it can implement the following steps to use a distributed computing framework to quickly classify and process key information with metadata tags to form a structured data set: Configure a distributed computing cluster, such as a Hadoop cluster or a Spark cluster; Install and configure the corresponding distributed computing framework software; Obtain key information with metadata tags from the company's internal systems (such as transaction systems, customer management systems, etc.) and external data sources (such as credit rating agencies, market research reports, etc.); Use the distributed computing framework to process the key information with metadata tags. Specifically, it can include the following steps: Initially group the data according to the metadata tags; Apply classification algorithms: Deploy machine learning algorithm models in the distributed computing framework and apply corresponding classification algorithms to the key information with different metadata tags; These algorithms may be trained based on historical data and can effectively distinguish different types of data; Perform secondary processing on the results of the initial grouping, merge adjacent or overlapping data groups, and ensure that all data within the groups belong to the same category through a consistency verification algorithm; Use data compression technology to compress and store the data after classification processing, reducing storage costs while retaining the integrity and availability of the data; Form a structured data set for subsequent data analysis, report generation, and other operations.

[0069] This application considers that in the financial services field, when dealing with tasks such as customer transaction records with large amounts of data, the traditional single-machine processing method faces performance bottlenecks, especially when dealing with real-time data streams, where the data volume is huge and rapid response is required. In addition, directly using unstructured raw data for analysis is often inefficient and difficult to meet business requirements. Therefore, an efficient data processing method is needed that can use a distributed computing framework to quickly classify and process key information with metadata tags to form a structured data set, thereby improving the efficiency and accuracy of data processing.

[0070] To solve the above problems, an optional solution is proposed in the embodiments of the present invention. By using a distributed computing framework to quickly classify and process key information with metadata tags to form a structured data set, the efficiency and accuracy of data processing are improved.

[0071] The optional solution is as follows:

[0072] Optionally, the "using a distributed computing framework to quickly classify and process key information with metadata tags to form a structured data set" described in step 104 includes:

[0073] Using a distributed computing framework, preliminarily group the key information according to metadata tags to obtain the result of preliminary grouping; deploy a machine learning algorithm model in the distributed computing framework, and apply corresponding classification algorithms to the key information with different metadata tags. The classification algorithms are trained based on historical data and can effectively distinguish different types of data; perform secondary processing on the result of the preliminary grouping, merge adjacent or overlapping data groups, and ensure that the data within all groups belongs to the same category through a consistency verification algorithm; use data compression technology to compress and store the classified data, reducing the storage cost while retaining the integrity and availability of the data to form a structured data set.

[0074] Suppose a financial services company wants to improve its risk management capabilities and better understand customer behavior. The specific implementation steps are as follows:

[0075] Configure an Apache Spark cluster to ensure there are enough computing nodes to process large-scale data; obtain key information with metadata tags from the company's internal systems (such as transaction systems, customer management systems, etc.), such as transaction records, customer profiles, etc.; obtain key information with metadata tags from external data sources (such as credit rating agencies, market research reports, etc.), such as credit scores, market trends, etc.; use Spark to preliminarily group the key information with metadata tags according to the tags. For example, group transaction records according to transaction types (deposit, withdrawal, transfer, etc.); deploy machine learning models in Spark, such as decision trees, random forests, etc., and apply corresponding classification algorithms to the key information with different metadata tags. For example, use a decision tree model to identify high-risk transactions; perform secondary processing on the result of the preliminary grouping, merge adjacent or overlapping data groups. For example, merge multiple small deposit records into one larger deposit record; use a consistency verification algorithm to ensure the consistency of the data within the group. For example, ensure that all transactions marked as "high-risk" actually meet the high-risk criteria; use data compression technology (such as GZIP, Snappy, etc.) to compress and store the processed data, reducing the storage cost; ensure that the compressed data still has complete availability, and finally form a structured data set.

[0076] 105. Provide an access interface to the structured data set to authorized users through a permission management system, enabling users to implement customized data analysis requests based on the interface;

[0077] This step refers to controlling users' access rights to structured data sets through a permission management system and providing one or more APIs (Application Programming Interfaces) or other types of interfaces, enabling authorized users to submit customized data analysis requests according to their own needs. This can ensure that only authenticated and authorized users can access sensitive data and can provide different data access and services according to different user roles and requirements.

[0078] Suppose a retail chain company hopes to improve its data-driven decision-making ability and ensure data security and compliance. Then it can implement the following steps to achieve this goal: Have the highest level of permissions to access all data and manage the permissions of other users; Can access data within a specific range and perform data analysis tasks; Can only access limited data related to their responsibilities and can only view specific reports or statistical data; Verify user identities through methods such as username and password, two-factor authentication, etc.; Assign different access permissions according to users' roles to ensure that users can only access data within their permission scope; Provide RESTful API interfaces for users to send data analysis requests through HTTP requests and receive response data; Provide a graphical interface for users without programming skills so that they can submit data analysis requests through simple click operations; Users can specify parameters such as the required data dimensions and time ranges through the API or graphical interface and submit data analysis requests; The system extracts relevant data from the structured data set according to the request content and returns the analysis results.

[0079] 106. In response to the data analysis request, dynamically adjust the presentation form of the structured data set to support the display of data views in different dimensions required during the decision-making process.

[0080] This step means that when users submit data analysis requests through the previous interfaces, the system needs to be able to dynamically adjust the display form of the structured data set according to the users' requests. This means that the system not only needs to be able to quickly process data analysis requests but also needs to be able to provide data view displays in multiple dimensions according to the needs and perspectives of different users, thus supporting more effective decision-making.

[0081] Suppose a logistics company hopes to optimize its transportation routes and cargo delivery efficiency. Then it can implement the following steps to respond to the data analysis request and dynamically adjust the data display form:

[0082] The user submits a data analysis request through the front - end interface or API interface, and the request contains the data dimension information required by the user. For example, the user may need to view the goods delivery situation in a specific region in the past month; the system extracts relevant data subsets from the structured data set according to the data dimension information in the request; applies data aggregation algorithms to aggregate the data subsets to generate a data summary result suitable for display. For example, calculates indicators such as the average daily delivery times, delay rate, etc.; according to the data dimension information and analysis requirements selected by the user, the system dynamically adjusts the data display form. For example, if the user needs to view the change trend of the daily delivery times, the system will generate a line chart to display this trend; if the user needs to view the distribution of the delivery delay rate within a specific time period, the system will generate a bar chart or heat map to display this distribution; uses dynamic data visualization components to display the data summary result; for example, uses visualization libraries such as ECharts, D3.js, etc. to generate charts; the user can view and interact with these charts in real - time through the front - end interface to deeply analyze the data and make decisions.

[0083] This application considers that in the logistics industry, managers need to optimize transportation routes and improve the efficiency of goods delivery based on real - time data. However, the existing data display methods are often fixed and cannot be dynamically adjusted according to the specific needs and perspectives of users, resulting in inflexible data display and affecting the efficiency and accuracy of decision - making. In addition, due to the large volume and complexity of data, it is difficult to quickly draw conclusions simply by relying on manual analysis. Therefore, a system that can dynamically adjust the data display form according to the user's request is needed to support the display of different - dimensional data views required in the decision - making process.

[0084] To solve the above problems, an optional solution is proposed in the embodiments of the present invention. By receiving the user's request and dynamically adjusting the data display form according to the request, it supports the display of different - dimensional data views, thereby improving the flexibility of data display and the efficiency of decision - making.

[0085] The optional solution is as follows:

[0086] Optionally, responding to the data analysis request and dynamically adjusting the presentation form of the structured data set to support the display of different - dimensional data views required in the decision - making process in step 106 includes:

[0087] Receive the data analysis request submitted by the user through the front-end interface or API interface, where the data analysis request contains the data dimension information required by the user; extract the relevant data subset from the structured dataset according to the data dimension information in the data analysis request, and the data subset is filtered from the structured dataset according to the data dimension information selected by the user; apply the data aggregation algorithm to aggregate the data subset to generate a data summary result suitable for display; use the dynamic data visualization component to display the data summary result.

[0088] Suppose a logistics company hopes to optimize its transportation routes and cargo delivery efficiency, and the specific implementation steps are as follows: The user submits a data analysis request through the front-end interface, and the request contains the data dimension information required for viewing. For example, the user needs to view the cargo delivery situation in the Beijing area in the past month, including indicators such as the number of daily deliveries and the delay rate; the system extracts the relevant data subset from the structured dataset according to the data dimension information in the request (such as time, location, number of deliveries, etc.). For example, filter out all delivery records in the Beijing area in the past month; apply the data aggregation algorithm to process the data subset to generate a data summary result suitable for display. For example, calculate indicators such as the average number of daily deliveries and the delay rate.

[0089] Use statistical formulas to calculate the average number of deliveries:

[0090]

[0091] Suppose the total number of deliveries in the Beijing area in the past month is 3,000 times, and the total number of days is 30 days, then the average number of deliveries is:

[0092]

[0093] According to the data dimension information and analysis requirements selected by the user, the system dynamically adjusts the data display form. For example, if the user needs to view the change trend of the number of daily deliveries, the system will generate a line chart to display this trend; if the user needs to view the distribution of delivery delay rates within a specific time period, the system will generate a bar chart or heat map to display this distribution; use ECharts to generate line charts and bar charts to display the data summary result. For example, generate a line chart of the number of daily deliveries in the past month and generate a bar chart of the daily delay rate.

[0094] This application takes into account that in a big data environment, traditional data aggregation methods often rely on fixed rules or simple statistical methods, which are insufficient when faced with complex and ever-changing data environments. Especially in the logistics industry, the spatio-temporal characteristics of data are obvious, and a single data aggregation method is difficult to comprehensively reflect the true situation of the data, resulting in decision-makers having difficulty obtaining accurate information to assist in decision-making. In addition, with the popularization of Internet of Things devices, real-time data streams have become increasingly important. How to effectively process these data streams and extract valuable information from them has become an urgent problem to be solved.

[0095] To solve the above problems, an optional solution for data aggregation with dynamic weights and spatio-temporal awareness based on a deep reinforcement learning model with multi-modal fusion is proposed in the embodiments of the present invention, aiming to more accurately capture the temporal and spatial characteristics of data, while considering factors such as context, prediction, reward mechanism, and behavior pattern, so as to provide more accurate data aggregation results.

[0096] The optional solution is as follows:

[0097] Optionally, the application data aggregation algorithm aggregates the data subset to generate a data summary result suitable for display, including:

[0098] For a given data subset D = {d1, d2, …, d n}, each data point d i contains multiple attributes a ij , where j represents the jth attribute. The data aggregation algorithm is defined as a data aggregation formula with dynamic weights and spatio-temporal awareness based on a deep reinforcement learning model with multi-modal fusion:

[0099]

[0100] Among them, P(D, A) represents the data aggregation result of the data subset D on the attribute set A; W j (d i , t) is the dynamic weight of the jth attribute of the data point d i at time t, and the dynamic weight is dynamically adjusted through a deep reinforcement learning model with multi-modal fusion; f(a ij ) is a function that calculates the corresponding aggregation value according to the type of the attribute a ij . α is the influence coefficient of the context factor, and the influence coefficient is dynamically adjusted according to the importance of the current context through an adaptive algorithm. G(Context(d i ), t) is a function that, according to the data point d iAdjust the final aggregation result according to the context information of i and its timestamp t. The function also learns the correlation between context and data points through a deep reinforcement learning model; γ is the influence coefficient of the spatial location factor, and this coefficient is also dynamically adjusted through an adaptive algorithm; H(Spatial(d i ),t) is a function that further adjusts the data aggregation result according to the spatial location information of data point d ij and its timestamp t. The function learns the correlation between geographical location and data through a deep reinforcement learning model; δ is the influence coefficient of the prediction result of the machine learning model; M(a ij ,t) is a prediction function based on time series analysis, used to predict the trend of a certain attribute a i in the future time period. The function can use a deep learning model for prediction; θ is the influence coefficient of the reward mechanism; R(d i ,t) is a reward function that determines the contribution of data point d i to the final aggregation result according to its importance at time t. The function is dynamically adjusted by a deep reinforcement learning model according to historical behavior; λ is the behavior influence coefficient; B(d i ,t) is a behavior function that adjusts the influence on the data aggregation result according to the behavior pattern associated with data point d

[0101] f(a ij ): This is a function for attribute a ij . It may be a normalization or transformation process of the attribute value, so that different attribute values can be compared on the same scale. A simple example is the normalization function, which maps all attribute values to between 0 and 1.

[0102]

[0103] where A is the set of all possible values of attribute a; G(Context(d i ),t) means this is a function adjusted according to the context information Context(d i ) and time t, and may be used to evaluate the importance of a certain data point d i under specific time and context conditions.

[0104]

[0105] Among them, H(Spatial(d i ),t) means this is a function adjusted according to the spatial location information Spatial(d i ) and time t, and may be used to evaluate the importance of a certain data point d iThe change in importance at a specific geographical location. For example, for a distribution point, a higher weight may be required during peak traffic hours.

[0106]

[0107] Among them, M(a ij , t) refers to a metric function related to attribute a ij and time t, which may be used to measure the trend of attribute values over time.

[0108]

[0109] B(d i , t): This is a behavior function based on data point d i and time t, which may be used to capture the behavior pattern of the data point over a period of time in the past.

[0110]

[0111] Suppose a logistics company wants to analyze its cargo distribution situation in the Beijing area in the past month, including indicators such as the average number of daily deliveries and the delay rate. The company hopes to see the change trend of the daily delivery times and understand which areas have a high delay rate: extract all distribution records in the Beijing area in the past month from the database; use the above data aggregation algorithm to process the data subset to generate a data summary result suitable for display; use ECharts to generate a line chart to show the change trend of the daily delivery times, and generate a bar chart or heat map to show the distribution of the delivery delay rate within a specific time period.

[0112] This application takes into account that existing data visualization methods are usually fixed and cannot dynamically adjust the chart display form according to the specific needs and scenarios of users. This may lead to non-intuitive data display or an incomplete reflection of the true situation of the data in practical applications, especially in fields such as the logistics industry that require quick response and decision-making. Therefore, an alternative solution that can dynamically adjust data visualization according to multiple factors such as user preferences, context information, spatial location, time context, historical behavior patterns, etc. is needed to improve the flexibility and pertinence of data display, and thus enhance the accuracy and efficiency of decision-making.

[0113] The alternative solution is as follows:

[0114] Optionally, displaying the data summary result using a dynamic data visualization component includes:

[0115] Generating a visualization chart according to the data summary result P(D, A), and the visualization chart is used to reflect the change trend of the data over time, including line charts, bar charts, and heat maps;

[0116] The generation algorithm of the visualization chart for the dynamic data is defined as:

[0117]

[0118] Among them, V(P) represents the visualization chart of the data aggregation result P(D,A), is the visualization weight adjusted according to the user preference U; C i (C) is the context weight adjusted according to the context information C; S i (S) is the position weight adjusted according to the spatial location information S; T context (T context ) is the time context weight adjusted according to the context time T context ; H i (H) is the historical behavior weight adjusted according to the historical behavior pattern H; U(U) is a function that selects the appropriate chart type according to the user preference U; T(T) is a function that automatically adjusts the time axis display of the chart according to the time dimension T.

[0119] Suppose a logistics company wants to optimize the efficiency of its cargo distribution in the Beijing area. The user hopes to view the cargo distribution situation in the Beijing area in the past month through the system, especially paying attention to the number of daily deliveries and the delay rate: The user submits a data analysis request through the front-end interface, requesting to view the cargo distribution situation in the Beijing area in the past month; The system extracts the relevant data subset from the structured data set according to the request; The data subset is processed by the data aggregation algorithm to generate a data aggregation result, such as indicators such as the average number of daily deliveries and the delay rate; The dynamic data visualization component is used to generate a visualization chart to display the data aggregation result.

[0120] Suppose the user preference is to view the change trend of the number of daily deliveries and prefers a bar chart display, while the context information indicates that there have been many delivery delays in the recent week, the location information shows that the delivery volume in Haidian District is the largest, the time context indicates that the number of deliveries on weekends is low, and the historical behavior pattern shows that the user often views the delay rate.

[0121] According to the above information, we can calculate the visualization chart V(P):

[0122]

[0123] Suppose the maximum value of P(d i ,a ij ) is 100, the user preference weight V i (U) is 0.8, the context weight C i (C) is 0.6, the position weight S i (S) is 0.7, the time context weight T context(T context ) is 0.5, and the historical behavior weight H i (H) is 0.4. The user preference function U(U) selects a bar chart, and the time dimension function T(T) is set to display by day. Then the calculation result of the visualization chart V(P) is:

[0124]

[0125] Figure 2 This is a schematic structural diagram of an XX device (or system) provided by an embodiment of the present application, as Figure 2 shown. The device includes:

[0126] A receiving module 21, configured to receive real-time data streams from an enterprise internal information system and an external public data source;

[0127] An identification module 22, configured to automatically identify key information in the real-time data stream according to preset business rules and map it to corresponding activity categories;

[0128] An integration module 23, configured to integrate the key information into a unified database and generate corresponding metadata tags for each category of key information;

[0129] A classification module 24, configured to perform fast classification processing on the key information with metadata tags by using a distributed computing framework to form a structured data set;

[0130] A providing module 25, configured to provide an access interface to the structured data set to authorized users through a permission management system, so that users can implement customized data analysis requests based on the interface;

[0131] An adjustment module 26, configured to dynamically adjust the presentation form of the structured data set in response to the data analysis request to support different-dimensional data view displays required in the decision-making process.

[0132] Figure 2 The described big data classification and utilization device for business management can execute Figure 1 the big data classification and utilization method for business management described in the embodiment shown, and its implementation principle and technical effects will not be elaborated. For the big data classification and utilization device for business management in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0133] In a possible design, Figure 2 the big data classification and utilization device for business management in the embodiment shown can be implemented as a computing device, as Figure 3As shown, the computing device may include a storage component 31 and a processing component 32;

[0134] The storage component 31 stores one or more computer instructions, and among them, the one or more computer instructions are for the processing component 32 to call and execute.

[0135] The processing component 32 is used to: receive real-time data streams from the enterprise internal information system and external public data sources; automatically identify key information in the real-time data streams according to preset business rules and map it to corresponding activity categories; integrate the key information into a unified database and generate corresponding metadata tags for each category of key information; use a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set; provide an access interface to the structured data set to authorized users through a permission management system; and dynamically adjust the presentation form of the structured data set in response to the data analysis request.

[0136] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0137] The storage component 31 is configured to store various types of data to support operations on the terminal. The storage component may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0138] Of course, the computing device may also necessarily include other components, such as input / output interfaces, display components, communication components, etc.

[0139] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module may be an output device, an input device, etc.

[0140] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0141] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server. The above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from a cloud computing platform.

[0142] An embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above-mentioned Figure 1 A method for classifying and utilizing big data for business management shown in the embodiment.

[0143] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0144] The device embodiments described above are merely illustrative. 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 to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0145] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A method for classifying and utilizing big data for business management, characterized in that, Including: Receiving real-time data streams from internal enterprise information systems and external public data sources; Automatically identifying key information in the real-time data streams according to preset business rules and mapping it to corresponding activity categories; Integrating the key information into a unified database and generating corresponding metadata tags for each category of key information; Using a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set; Providing an access interface to the structured data set to authorized users through a permission management system, enabling users to implement customized data analysis requests based on the interface; Responding to the data analysis request and dynamically adjusting the presentation form of the structured data set to support the display of data views in different dimensions required during the decision-making process.

2. The method according to claim 1, wherein The automatically identifying key information in the real-time data streams according to preset business rules and mapping it to corresponding activity categories includes: Parsing the text content in the real-time data streams through natural language processing technology to identify first key information containing specific keywords or phrases; Using predefined regular expressions or pattern matching algorithms to screen structured or semi-structured data items in the real-time data streams as second key information, where the regular expressions or pattern matching algorithms are customized based on preset business rules to capture data items conforming to specific formats; Based on the preset business rules, applying a machine learning model to analyze the real-time data streams to obtain third key information, where the machine learning model is constructed through learning a large amount of historical data to identify potential and non-explicit key information; Taking the first key information, the second key information, and the third key information as key information and mapping them to preset activity categories.

3. The method according to claim 1, wherein The integrating the key information into a unified database and generating corresponding metadata tags for each category of key information includes: Importing the identified and classified key information into a unified database; Creating corresponding data tables or partitions for each category of key information in the database; Using a metadata management system to assign unique identifiers to the key information and generating corresponding metadata tags according to the attributes of the key information, where the attributes at least include source, type, creation time, and update time.

4. The method according to claim 1, wherein The using a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set includes: Using a distributed computing framework to preliminarily group the key information according to metadata tags to obtain the result of preliminary grouping; Deploying a machine learning algorithm model in the distributed computing framework and applying corresponding classification algorithms to the key information with different metadata tags, where the classification algorithms are trained based on historical data and can effectively distinguish different types of data; Performing secondary processing on the result of preliminary grouping, merging adjacent or overlapping data groups, and ensuring that the data within all groups belongs to the same category through a consistency verification algorithm. Use data compression technology to compress and store the classified data, reducing storage costs while retaining the integrity and availability of the data to form a structured data set.

5. The method according to claim 1, wherein In response to the data analysis request, dynamically adjust the presentation form of the structured data set to support the display of different-dimensional data views required in the decision-making process, including: Receive a data analysis request submitted by the user through the front-end interface or API interface, where the data analysis request contains the data dimension information required by the user; According to the data dimension information in the data analysis request, extract relevant data subsets from the structured data set, and the data subsets are filtered from the structured data set according to the data dimension information selected by the user; Apply a data aggregation algorithm to aggregate the data subsets to generate a data summary result suitable for display; Use a dynamic data visualization component to display the data summary result.

6. The method according to claim 5, characterized in that The application of the data aggregation algorithm to aggregate the data subsets to generate a data summary result suitable for display includes: For a given data subset D = {d1, d2, …, d n}, each data point d i contains multiple attributes a ij , where j represents the j-th attribute, and the data aggregation algorithm is defined as a dynamic weight and spatio-temporal awareness data aggregation formula based on a multi-modal fusion deep reinforcement learning model: Among them, P(D, A) represents the data aggregation result of the data subset D on the attribute set A; W j (d i , t) is the dynamic weight of the j-th attribute of the data point d i at time t, and the dynamic weight is dynamically adjusted by a deep reinforcement learning model based on multimodal fusion; f(a ij ) is a function that calculates the corresponding aggregation value according to the type of the attribute a ij , α is the influence coefficient of the context factor, and the influence coefficient is dynamically adjusted by an adaptive algorithm according to the importance of the current context. G(Context(d i ), t) is a function that adjusts the final aggregation result according to the context information of the data point d i and its timestamp t. The function also learns the correlation between the context and the data point through a deep reinforcement learning model; γ is the influence coefficient of the spatial location factor, and this coefficient is also dynamically adjusted by an adaptive algorithm; H(Spatial(d i ), t) is a function that further adjusts the data aggregation result according to the spatial location information of the data point d i and its timestamp t. The function learns the correlation between the geographical location and the data through a deep reinforcement learning model; δ is the influence coefficient of the prediction result of the machine learning model; M(a ij , t) is a prediction function based on time series analysis, which is used to predict the trend of a certain attribute a ij in the future time period. The function can use a deep learning model for prediction; θ is the influence coefficient of the reward mechanism; R(d i , t) is a reward function that determines the contribution of the data point d i at time t to the final aggregation result according to its importance. The function is dynamically adjusted by a deep reinforcement learning model according to historical behaviors; λ is the behavior influence coefficient; B(d i , t) is a behavior function that adjusts the influence on the data aggregation result according to the behavior pattern associated with the data point d i .

7. The method according to claim 5, characterized in that, The use of a dynamic data visualization component to display the data summary result includes: Generate a visualization chart according to the data summary result P(D,A), and the visualization chart is used to reflect the change trend of the data over time, including line charts, bar charts, and heat maps; The generation algorithm of the visualization chart of the dynamic data is defined as: Among them, V(P) represents the visualization chart of the data aggregation result P(D,A), and v i (U) is the visualization weight adjusted according to the user preference U; C i (C) is the context weight adjusted according to the context information C; S i (S) is the position weight adjusted according to the spatial location information S; T context (T context ) is the time context weight adjusted according to the context time T context ; H i (H) is the historical behavior weight adjusted according to the historical behavior pattern H; U(U) is a function that selects the appropriate chart type according to the user preference U; T(T) is a function that automatically adjusts the time axis display of the chart according to the time dimension T.

8. A big data classification and utilization system for business management, characterized in that, Include: A receiving module for receiving real-time data streams from enterprise internal information systems and external public data sources; An identification module for automatically identifying key information in the real-time data stream according to preset business rules and mapping it to corresponding activity categories; An integration module for integrating the key information into a unified database and generating corresponding metadata tags for each category of key information; A classification module for using a distributed computing framework to quickly classify the key information with metadata tags to form a structured data set; A providing module for providing an access interface to the structured data set to authorized users through a permission management system, enabling users to implement customized data analysis requests based on the interface; An adjustment module for dynamically adjusting the presentation form of the structured data set in response to the data analysis request to support the display of different-dimensional data views required in the decision-making process.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method and system for classifying and utilizing big data for business management as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, A computer program is stored, and when the computer program is executed by a computer, it implements a method and system for classifying and utilizing big data for business management as described in any one of claims 1 to 7.

Images