Enterprise service data intelligent management system based on cloud platform

Abstract

The present invention relates to the technical field of enterprise service information management, and in particular to an enterprise service data intelligent management system based on a cloud platform, comprising extracting key information elements from user service data, recombining them according to semantic information, obtaining the information body of the service data, then arranging the positions of the service data of the same user in different time periods, calculating the information correlation values ​​between adjacent service nodes, integrating the service data in different time periods, obtaining real-time standard data, enterprise market data and matching sales data in each cycle time, calculating the data application value of the real-time standard data, and calculating the data storage cycle of the real-time standard data based on the data application value and the shortest storage cycle of the real-time standard data. After that, a cloud storage module stores the real-time standard data according to the data storage cycle, thereby optimizing the resource allocation of storage resources and effectively utilizing the storage space.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise service information management, and in particular to an enterprise service data intelligent management system based on a cloud platform. Background Art

[0002] Enterprise service data not only covers the internal management and operation information of the enterprise, but also includes market data, customer data and competition data related to enterprise services. These data are important bases for measuring enterprise service quality, optimizing service processes and improving customer satisfaction.

[0003] Prior art CN112418688A discloses an enterprise data management method based on an enterprise service portal, the method comprising the following steps: setting the fields of enterprise data, specifying the classification standards of the fields of enterprise data, and constructing a framework for enterprise data management; collecting a list of enterprises according to the framework for enterprise data management; collecting enterprise data from a data source according to the collected list of enterprises, evaluating the collected enterprise data, and storing the evaluated enterprise data in an enterprise data management database; processing the enterprise and the enterprise data content for the stored enterprise data; and constructing an enterprise data management statistical analysis module according to the processed enterprise data;

[0004] However, when enterprise service data is managed, the storage period of the service data does not match the actual application value of the data, which leads to the retention time of enterprise service data being too long or too short. The former wastes storage resources and increases security risks, while the latter may result in the loss of critical data. Summary of the invention

[0005] The purpose of the present invention is to solve the problems in the background technology and to propose an enterprise service data intelligent management system based on a cloud platform.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The enterprise service data intelligent management system based on the cloud platform includes data collection module, cloud storage module, application analysis module, data analysis module, data retrieval module and periodic operation module:

[0008] The data collection module is used to collect the enterprise's metadata and transmit it to the data analysis module. The metadata includes enterprise service data and enterprise market data;

[0009] Cloud storage module, used to store enterprise information;

[0010] The data retrieval module is used to identify the user of the real-time service data and mark the user as the service object. According to the identity information of the service object, the historical service data of the service object is retrieved and transmitted to the data analysis module;

[0011] The data analysis module is used to mark the real-time service data of the service object as a service node, and extract the information body of the service node based on the part of speech and semantics of the service data, then connect the real-time service data with the historical service data in chronological order, take the service node corresponding to the real-time service data as the real-time node, calculate the information correlation value between the real-time node and the front-end node, and determine the real-time standard data based on the information correlation value;

[0012] The application analysis module calculates the matching data ratio within the cycle time according to the matching sales data and the enterprise market data, and uses the formula to comprehensively calculate the data ratios within multiple cycle times to obtain the data application value of the real-time standard data, where the matching sales data refers to the sales volume corresponding to the historical matching information, and the enterprise market data refers to the market sales data of the enterprise products. Then, the application analysis module transmits the data application value to the cycle operation module;

[0013] The cycle calculation module calculates the data storage cycle of the real-time standard data according to the data application value and the shortest data storage time of the real-time standard data.

[0014] As a further solution of the present invention, a method for extracting the information body of a service node includes:

[0015] Each real-time service data is marked as a separate service node. In this case, the historical service data will contain several service nodes.

[0016] Select any service data of a service node as target data, segment the target data according to word meaning, obtain several data phrases, classify the data phrases according to their parts of speech, and obtain a data segment set, wherein the data phrases are elements in the data segment set;

[0017] Count the number of occurrences of each data phrase in the segment set, and divide the number of occurrences of the data phrase by the total number of data phrases in the segment set to obtain the phrase occurrence ratio of each data phrase in the segment set;

[0018] Taking the data segment set as a unit, comparing the phrase appearance ratios of the data phrases in each data segment set, and marking the data phrase corresponding to the maximum phrase appearance ratio as the key information element of this data set;

[0019] Summarize the key information elements in all the segment sets, use the service data of this service node as the semantic background, semantically combine the key information elements, and mark the combined result as the information body.

[0020] As a further solution of the present invention, a method for calculating the information relevance value includes:

[0021] Connect the real-time service data with the historical service data, arrange the service nodes in the real-time service data and the historical service data in chronological order, and mark the data segment after the arrangement as the service information segment;

[0022] The service information segment is converted into a numerical vector using the TF-IDF model algorithm. The service node corresponding to the real-time service data is used as a benchmark, and the service node corresponding to the real-time service data is marked as a real-time node. At the same time, a service node is obtained forward and marked as a front-end node.

[0023] Then, the Pearson correlation coefficient algorithm is used to calculate the information correlation value of the real-time node and the front-end node.

[0024] As a further solution of the present invention, the method for calculating the real-time standard data includes:

[0025] The information correlation value is compared with the correlation threshold X1. If the information correlation value is less than the correlation threshold X1, the real-time node is directly marked as real-time standard data. Otherwise, if the information correlation value is greater than or equal to the correlation threshold X1, the front-end node and the real-time node are merged into a comprehensive node. At the same time, a service node is obtained and repeatedly marked as a front-end node. Then, the information correlation value between the front-end node and the comprehensive node is calculated according to the above method, and the above steps are repeated in sequence until the real-time standard data is obtained.

[0026] As a further solution of the present invention, if the service object is a new user of the enterprise, the historical service data of the service object will not exist. At this time, the real-time service data of the service object is directly marked as real-time standard data.

[0027] As a further solution of the present invention, when real-time standard data is obtained, the data analysis module transmits the real-time standard data to the data retrieval module. The data retrieval module searches the cloud storage module according to the real-time standard data, and obtains historical matching data based on the retrieval results. Then, the data retrieval module transmits the historical matching data to the application analysis module.

[0028] As a further solution of the present invention, a method for calculating the data application value of real-time standard data includes:

[0029] SS1: Set the cycle time and divide the enterprise market data and matching sales data according to the cycle time;

[0030] SS2: Taking the current time as the time node, obtain the enterprise market data and matching sales data within the previous four cycles, and arbitrarily select a cycle time as the target cycle, and mark the enterprise market data within the target cycle as target market data DSi, where i represents the cycle time, and i∈[1,I], I takes the value of 4, indicating that there are 4 cycles;

[0031] SS3: Utilizing formulas Get the matching data ratio Bi of target period i, where DXi is the sum of matching sales data in the target period;

[0032] SS4: Based on the formula again The data application value Vf of the real-time standard data is obtained, where b = i-1, Bmax is the maximum value of the matching data ratio Bi, Bmin is the minimum value of the matching data ratio Bi, c is a fixed coefficient, and c>1, λ is a positive growth coefficient, that is, λ refers to In the operation, (Bi-Bb)≥0, β1 and β2 are both proportional coefficients.

[0033] As a further solution of the present invention, a method for calculating a data storage period includes:

[0034] Based on the formula CL = d Vf×θ1 The data storage period CL of the real-time standard data is obtained, wherein d is the shortest storage time of the real-time standard data, and d>1, and θ1 is the weight coefficient of the real-time standard data.

[0035] Compared with the prior art, the advantages of the present invention are:

[0036] The present invention extracts key information elements from the user's service data and recombines them according to the semantic information of the service data to obtain the information body of the service data, then arranges the service data of the same user in different time periods in chronological order, calculates the information correlation value between adjacent service nodes, integrates the service data of different time periods to obtain real-time standard data, and stores the real-time standard data, thereby centrally managing the user's service information and facilitating the implementation of unified security measures. At the same time, the enterprise can allocate resources more efficiently and improve the efficiency and quality of the service process;

[0037] The present invention calculates the data application value of real-time standard data according to the enterprise market data and matching sales data within each cycle time, and calculates the data storage period of real-time standard data based on the data application value and the shortest storage period of real-time standard data. Then the cloud storage module stores the real-time standard data according to the data storage period. On the one hand, it optimizes the resource allocation of storage resources and effectively utilizes the storage space. On the other hand, it reduces the security risks caused by long-term data storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0040] Reference Figure 1 , an enterprise service data intelligent management system based on a cloud platform, including a data acquisition module, a cloud storage module, an application analysis module, a data analysis module, a data retrieval module and a periodic operation module;

[0041] The data collection module is used to collect the metadata of the enterprise, where metadata refers to data related to the enterprise, including enterprise service data, enterprise market data and financial data, etc. The enterprise data collection module and the data analysis module are connected in a one-way manner;

[0042] The cloud storage module is used to store enterprise information, and the cloud storage module is bidirectionally connected to the data retrieval module and the periodic operation module;

[0043] The data retrieval module is used to identify the real-time enterprise service data and mark it as real-time service data, and mark the users of the enterprise real-time service as service objects, and retrieve the historical service data of the service objects in the cloud storage module based on the identity information of the service objects, wherein the identity information includes the user name and ID number of the service objects;

[0044] Afterwards, the data retrieval module transmits the real-time service data and historical service data of the service object to the data analysis module respectively;

[0045] The data analysis module is used to process the real-time service data and historical service data of the service object and generate real-time standard data. The specific processing methods include:

[0046] S1: Mark each real-time service data as a separate service node. At this time, the historical service data will contain several service nodes.

[0047] S2: Identify the information subject of each service node, where the information subject refers to the key information in a single service node. The specific method for identifying the information subject includes:

[0048] S21: arbitrarily select service data of a service node as target data, segment the target data according to word meanings to obtain a number of data phrases, and then classify the data phrases according to the parts of speech of the data phrases to obtain a number of segment sets;

[0049] It should be further explained that the data phrase is an element in the data segment set. In this embodiment, the target data is segmented using a natural language processing model to divide the target data according to semantics. At the same time, when the data phrase is divided according to parts of speech, its parts of speech include nouns, verbs, etc.

[0050] S22: Count the number of occurrences of each data phrase in the number segment set, and divide the number of occurrences of the data phrase by the total number of data phrases in the number segment set to obtain the phrase occurrence ratio of each data phrase in the number segment set;

[0051] S23: Taking the data segment set as a unit, comparing the phrase appearance ratios of the data phrases in each data segment set, and marking the data phrase corresponding to the maximum phrase appearance ratio as a key information element of the data set;

[0052] S24: Summarize the key information elements in all the data segment sets, use the service data of the service node as the semantic background, semantically combine the key information elements, and mark the combined result as the information body;

[0053] The semantic combination of key information elements adopts natural language processing technology, and natural language processing technology is an existing technology, which will not be described in detail here;

[0054] S3: joining the real-time service data with the historical service data, arranging the service nodes in the real-time service data and the historical service data in chronological order, and marking the data segment after the arrangement as a service information segment;

[0055] S4: The service information segment is converted into a numerical vector using the TF-IDF model algorithm. The service node corresponding to the real-time service data is used as a benchmark, and the service node corresponding to the real-time service data is marked as a real-time node. At the same time, a service node is obtained forward and marked as a front-end node.

[0056] Then, the Pearson correlation coefficient algorithm is used to calculate the information correlation value of the real-time node and the front-end node, and the information correlation value is compared with the correlation threshold X1. If the information correlation value is less than the correlation threshold X1, the real-time node is directly marked as real-time standard data. On the contrary, if the information correlation value is greater than or equal to the correlation threshold X1, the front-end node and the real-time node are merged into a comprehensive node, and a service node is obtained forward and repeatedly marked as a front-end node. Then, the information correlation value between the front-end node and the comprehensive node is calculated according to the above method, and the above steps are repeated in sequence until the real-time standard data is obtained;

[0057] Among them, the TF-IDF model algorithm and the Pearson correlation coefficient algorithm are both existing technologies and will not be described in detail here. The relevant threshold value X1 is obtained by those skilled in the art after big data calculation;

[0058] It should be further explained that if the service object is a new user of the enterprise, the historical service data of this service object will not exist. At this time, the real-time service data of the service object is directly marked as real-time standard data;

[0059] Afterwards, the data analysis module transmits the real-time standard data to the data retrieval module, which searches the cloud storage module according to the real-time standard data and obtains the historical matching data according to the search results. Afterwards, the data retrieval module transmits the historical matching data to the application analysis module;

[0060] It should be further explained that when the real-time standard data is retrieved in the enterprise database, the real-time standard data is used as an input value and retrieved in the enterprise database to obtain historical matching information. At the same time, based on the historical matching information, the sales of the products in each historical matching information are identified and marked as matching sales data;

[0061] For example, within the cycle time, there is a user whose service data is consistent with the real-time standard data, and the purchase amount of the enterprise product by this user is a1. At this time, a1 is the matching sales data;

[0062] The application analysis module is used to determine the data application value of the real-time standard data by matching the sales data with the enterprise market data, where the enterprise market data refers to the data of the enterprise's products sold in the market. The specific method for determining the data application value of the real-time standard data includes:

[0063] SS1: Set the cycle time, and divide the enterprise market data and matching sales data according to the cycle time. In this embodiment, the cycle time is set to 3 months;

[0064] SS2: Taking the current time as the time node, obtain the enterprise market data and matching sales data within the previous four cycles, and arbitrarily select a cycle time as the target cycle, and mark the enterprise market data within the target cycle as target market data DSi, where i represents the cycle time, and i∈[1,I]. In this embodiment, I is 4, indicating that there are 4 cycles;

[0065] SS3: Utilizing formulas Get the matching data ratio Bi of target period i, where DXi is the sum of matching sales data in the target period;

[0066] Then, the remaining cycle time is taken as the target cycle in turn, and the matching data ratio Bi of each cycle time is calculated;

[0067] SS4: Then use the formula The data application value Vf of the real-time standard data is obtained, where b = i-1, Bmax is the maximum value of the matching data ratio Bi, Bmin is the minimum value of the matching data ratio Bi, c is a fixed coefficient, and c>1, λ is a positive growth coefficient, that is, In the calculation, the number of (Bi-Bb) ≥ 0, β1 and β2 are both proportional coefficients, and the specific values ​​of β1 and β2 are obtained by technicians in this field after big data calculation;

[0068] Then the application analysis module transmits the data application value Vf of the real-time standard data to the periodic operation module;

[0069] The cycle calculation module calculates the storage cycle of the real-time standard data based on the data application value of the real-time standard data to obtain the data storage cycle. The specific calculation method includes:

[0070] Based on the formula CL = d Vf×θ1 The data storage period of the real-time standard data is obtained, wherein d is the shortest storage time of the real-time standard data, further d>1, θ1 is the weight coefficient of the real-time standard data, wherein the shortest storage time d and θ1 are obtained by those skilled in the art after big data calculation;

[0071] Afterwards, the periodic operation module transmits the data storage period to the cloud storage module, and the cloud storage module stores the real-time standard data according to the data storage period.

[0072] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent management system for enterprise service data based on a cloud platform, characterized by: It includes data collection module, cloud storage module, application analysis module, data analysis module, data retrieval module and periodic operation module: The data collection module is used to collect the enterprise's metadata and transmit it to the data analysis module. The metadata includes enterprise service data and enterprise market data; Cloud storage module, used to store enterprise information; The data retrieval module is used to identify the user of the real-time service data and mark the user as the service object. According to the identity information of the service object, the historical service data of the service object is retrieved and transmitted to the data analysis module; The data analysis module is used to mark the real-time service data of the service object as a service node, and extract the information body of the service node based on the part of speech and semantics of the service data, then connect the real-time service data with the historical service data in chronological order, take the service node corresponding to the real-time service data as the real-time node, calculate the information correlation value between the real-time node and the front-end node, and determine the real-time standard data based on the information correlation value; Specifically, the calculation method of real-time standard data includes: Connect the real-time service data with the historical service data, arrange the service nodes in the real-time service data and the historical service data in chronological order, and mark the data segment after the arrangement as the service information segment; The service information segment is converted into a numerical vector using the TF-IDF model algorithm. The service node corresponding to the real-time service data is used as a benchmark, and the service node corresponding to the real-time service data is marked as a real-time node. At the same time, a service node is obtained forward and marked as a front-end node. Then, the Pearson correlation coefficient algorithm is used to calculate the information correlation value of the real-time node and the front-end node; Compare the information correlation value with the correlation threshold X1. If the information correlation value is less than the correlation threshold X1, the real-time node is directly marked as real-time standard data. Otherwise, if the information correlation value is greater than or equal to the correlation threshold X1, the front-end node and the real-time node are merged into a comprehensive node. At the same time, a service node is obtained and repeatedly marked as a front-end node. Then, the information correlation value between the front-end node and the comprehensive node is calculated according to the above method. The above steps are repeated in sequence until the real-time standard data is obtained. The application analysis module calculates the matching data ratio within the cycle time according to the matching sales data and the enterprise market data, and uses the formula to comprehensively calculate the data ratios within multiple cycle times to obtain the data application value of the real-time standard data, where the matching sales data refers to the sales volume corresponding to the historical matching information, and the enterprise market data refers to the market sales data of the enterprise products. After that, the application analysis module transmits the data application value to the cycle operation module. Specifically, the calculation method of the data application value includes: SS1: Set the cycle time and divide the enterprise market data and matching sales data according to the cycle time; SS2: Taking the current time as the time node, obtain the enterprise market data and matching sales data within the previous four cycles, and arbitrarily select a cycle time as the target cycle, and mark the enterprise market data within the target cycle as target market data DSi, where i represents the cycle time, and i∈[1,I], I takes the value of 4, indicating that there are 4 cycles; SS3: Utilizing formulas Get the matching data ratio Bi of target period i, where DXi is the sum of matching sales data in the target period; SS4: Based on the formula again The data application value Vf of the real-time standard data is obtained, where b = i-1, Bmax is the maximum value of the matching data ratio Bi, Bmin is the minimum value of the matching data ratio Bi, c is a fixed coefficient, and c>1, λ is a positive growth coefficient, that is, λ refers to In the operation, the number of (Bi-Bb) ≥ 0, β1 and β2 are both proportional coefficients; The cycle calculation module calculates the data storage cycle of the real-time standard data according to the data application value and the shortest data storage time of the real-time standard data.

2. The cloud platform-based enterprise service data intelligent management system according to claim 1, characterized in that: The method for extracting the information body of the service node includes: Each real-time service data is marked as a separate service node. In this case, the historical service data will contain several service nodes. Select any service data of a service node as target data, segment the target data according to word meaning, obtain several data phrases, classify the data phrases according to their parts of speech, and obtain a data segment set, wherein the data phrases are elements in the data segment set; Count the number of occurrences of each data phrase in the segment set, and divide the number of occurrences of the data phrase by the total number of data phrases in the segment set to obtain the phrase occurrence ratio of each data phrase in the segment set; Taking the data segment set as a unit, comparing the phrase appearance ratios of the data phrases in each data segment set, and marking the data phrase corresponding to the maximum phrase appearance ratio as the key information element of this data set; Summarize the key information elements in all the segment sets, use the service data of this service node as the semantic background, semantically combine the key information elements, and mark the combined result as the information body.

3. The cloud platform-based enterprise service data intelligent management system according to claim 1, characterized in that: If the service object is a new user of the enterprise, the historical service data of this service object will not exist. At this time, the real-time service data of the service object is directly marked as real-time standard data.

4. The cloud platform-based enterprise service data intelligent management system according to claim 1, characterized in that: When real-time standard data is obtained, the data analysis module transmits the real-time standard data to the data retrieval module. The data retrieval module searches the cloud storage module according to the real-time standard data, and obtains historical matching data based on the retrieval results. The data retrieval module then transmits the historical matching data to the application analysis module.

5. The cloud platform-based enterprise service data intelligent management system according to claim 1, characterized in that: The calculation method of data storage period includes: Based on the formula CL = d Vf×θ1 The data storage period CL of the real-time standard data is obtained, wherein d is the shortest storage time of the real-time standard data, and d>1, and θ1 is the weight coefficient of the real-time standard data.

Images