Enterprise data application management system and method

Abstract

The invention discloses an enterprise data application management system and method, and the system comprises an obtaining module which is used for capturing the public data of an enterprise and receiving the internal data of the enterprise; the screening and checking module is used for carrying out preprocessing of cleaning, duplicate removal and graphic conversion on the public data and checking default values of internal data to obtain enterprise target data; the analysis module is used for performing text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, a financial analysis result, a market status analysis result, advantage information, disadvantage information, opportunity information and risk information of the enterprise; and the generation module is used for calling a preset report template, adding the basic information, the financial analysis result, the market status analysis result, the advantage information, the disadvantage information, the opportunity information and the risk information into the preset report template, and generating and outputting a data application management report of the enterprise. According to the invention, information related to an enterprise is comprehensively and accurately acquired and efficiently and accurately analyzed.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an enterprise data application management system and method. Background Art

[0002] In the Internet era, data and information appear in massive amounts. For enterprises, in addition to paying attention to the data and information of the enterprise itself, the development of the enterprise also needs to pay attention to various types of information such as its industry, region, market economic environment, and policies. This type of information is also massive, and enterprise decision makers often find it difficult to obtain comprehensive and accurate information, which makes the results generated based on the analysis of the obtained information inaccurate. At the same time, there is also the problem of low analysis efficiency. Therefore, how to comprehensively and accurately obtain information related to the enterprise and conduct efficient and accurate analysis is a technical problem that needs to be solved at present. Summary of the invention

[0003] The main purpose of the present invention is to provide an enterprise data application management system and method, aiming to solve the technical problem of how to comprehensively and accurately obtain enterprise-related information and perform efficient and accurate analysis in the prior art.

[0004] To achieve the above object, the present invention provides an enterprise data application management system, the enterprise data application management system comprising:

[0005] An acquisition module is used to capture the public data of an enterprise and receive the internal data of an enterprise. The public data includes but is not limited to news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data includes but is not limited to employee information, financial data and market data;

[0006] A screening and verification module is used to perform cleaning, deduplication and graphic conversion preprocessing on the public data, and to perform default value verification on the internal data to obtain enterprise target data;

[0007] An analysis module, used to perform text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise;

[0008] The generation module is used to call a preset report template, and add the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template to generate and output the data application management report of the enterprise.

[0009] Preferably, the enterprise data application management system further includes:

[0010] A monitoring module, used for monitoring whether there is first updated data corresponding to the public data and whether there is second updated data corresponding to the internal data at every preset period;

[0011] an updating module, configured to update the data application management report when the first update data and / or the second update data exist;

[0012] The display module is used to control the display of the updated data application management report.

[0013] Preferably, the enterprise data application management system further includes:

[0014] A first judgment module is used to, when receiving an access request, judge whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report;

[0015] An encryption module is used to generate an encryption key based on a preset algorithm when the account authority matches the report authority, and encrypt the data application management report based on the encryption key, wherein the preset algorithm is:

[0016] X n =(K*X n-1 +N)mod W;

[0017] Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation;

[0018] A transmission module is used to transmit the encrypted data application management report to the access user account.

[0019] Preferably, the enterprise data application management system further includes:

[0020] A collection module, used to collect enterprise data samples, wherein the enterprise data samples include at least news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples;

[0021] A training module, used to train a preset network model according to the enterprise data sample and calculate the loss function value of the preset network model;

[0022] A second judgment module is used to judge whether the loss function value is less than a preset loss threshold, and if it is less than the preset loss threshold, the preset network model is generated as the preset semantic network model;

[0023] An updating module is used to update the weight value and bias value in the preset network model when the loss function value is not less than the preset loss threshold, wherein the updating formula is:

[0024]

[0025] Wherein, w represents the updated weight value, w0 represents the weight value before the update, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before the update;

[0026] Execution module: used for executing the step of training the preset network model according to the enterprise data sample for the updated preset network model.

[0027] Preferably, the screening and verification module comprises:

[0028] A search unit: used for searching the public data for noise data irrelevant to the enterprise, and removing the noise data;

[0029] A screening unit, used to determine duplicate data items in the public data, select any one of the duplicate data items as a reserved item, and delete other items of the duplicate data items except the reserved item;

[0030] The recognition unit is used to recognize the graphics in the public data and determine the text information in the graphics.

[0031] Furthermore, to achieve the above-mentioned purpose, the present invention also provides an enterprise data application management method, the enterprise data application management method comprising:

[0032] Capture the public data of enterprises and receive the internal data of enterprises. The public data includes but is not limited to news information, litigation information, public opinion information, industry reports and policy documents; the internal data includes but is not limited to employee information, financial data and market data;

[0033] Pre-processing the public data by cleaning, de-duplication and graphic conversion, and performing default value verification on the internal data to obtain enterprise target data;

[0034] Performing text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise;

[0035] The preset report template is called, and the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are added to the preset report template to generate and output the data application management report of the enterprise.

[0036] Preferably, the step of generating and outputting the data application management report of the enterprise includes:

[0037] At every preset period, monitoring whether there is first updated data corresponding to the public data, and whether there is second updated data corresponding to the internal data;

[0038] If the first update data and / or the second update data exist, updating the data application management report;

[0039] Controlling display of the updated data application management report.

[0040] Preferably, the step of generating and outputting the data application management report of the enterprise further includes:

[0041] When receiving an access request, determining whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report;

[0042] If the account authority matches the report authority, an encryption key is generated based on a preset algorithm, and the data application management report is encrypted based on the encryption key, wherein the preset algorithm is:

[0043] X n =(K*X n-1 +N)mod W;

[0044] Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation;

[0045] The encrypted data application management report is transmitted to the access user account.

[0046] Preferably, the step of performing text semantic analysis on the enterprise target data based on a preset semantic network model includes:

[0047] Collecting enterprise data samples, wherein the enterprise data samples include at least news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples;

[0048] Training a preset network model according to the enterprise data sample, and calculating a loss function value of the preset network model;

[0049] Determine whether the loss function value is less than a preset loss threshold, and if so, generate the preset network model as the preset semantic network model;

[0050] If the loss function value is not less than the preset loss threshold, the weight value and bias value in the preset network model are updated, wherein the update formula is:

[0051]

[0052] Wherein, w represents the updated weight value, w0 represents the weight value before the update, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before the update;

[0053] For the updated preset network model, a step of training the preset network model according to the enterprise data sample is performed.

[0054] Preferably, the steps of preprocessing the public data by cleaning, deduplication and graphic conversion include:

[0055] Searching for noise data in the public data that is irrelevant to the enterprise data, and removing the noise data;

[0056] Determine duplicate data items in the public data, select any one of the duplicate data items as a reserved item, and delete other items of the duplicate data items except the reserved item;

[0057] The graphics in the public data are identified to determine the text information in the graphics.

[0058] The enterprise data application management system and method of the present invention include an acquisition module, a screening and verification module, an analysis module and a generation module: wherein the acquisition module is used to capture the public data of the enterprise and receive the internal data of the enterprise, wherein the public data at least includes news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data at least includes employee information, financial data and market data; the screening and verification module is used to perform pre-processing of cleaning, de-duplication and graphic conversion of the public data, and perform default value verification on the internal data to obtain the enterprise target data; the analysis module is used to perform text semantic analysis on the enterprise target data based on a preset semantic network model to obtain the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise; the generation module is used to call a preset report template, and add the obtained basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template, generate the enterprise data application management report and output it. In this way, various public data related to the enterprise are collected with the help of crawler technology, making the collected information more comprehensive. At the same time, the uploaded internal data is combined to generate enterprise target data. The preset semantic network model formed by a large amount of data training is used to perform in-depth text semantic analysis to obtain the company's basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information, which are added to the preset report template to generate the company's data application management report, so as to achieve comprehensive and accurate acquisition of information related to the enterprise and conduct efficient and accurate analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic diagram of modules of the first embodiment of the enterprise data application management system of the present invention;

[0060] Figure 2 A module diagram of a second embodiment of the enterprise data application management system of the present invention;

[0061] Figure 3 It is a flowchart of the first embodiment of the enterprise data application management method of the present invention;

[0062] Figure 4 It is a flowchart of the second embodiment of the enterprise data application management method of the present invention.

[0063] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0064] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0065] The present invention provides an enterprise data application management system, please refer to Figure 1 , Figure 1 It is a module diagram of the first embodiment of the enterprise data application management system of the present invention.

[0066] Specifically, the enterprise data application management system in this embodiment includes an acquisition module 10, a screening and verification module 20, an analysis module 30 and a generation module 40, wherein:

[0067] The acquisition module 10 is used to capture the public data of the enterprise and receive the internal data of the enterprise. The public data includes but is not limited to news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data includes but is not limited to employee information, financial data and market data;

[0068] Furthermore, when an enterprise needs to perform data classification application management, it initiates a management request to the enterprise data application management system of this embodiment. After receiving the management request, the enterprise data application management system reads the enterprise ID (Identification) carried therein and determines the specific enterprise that needs data classification application management by the enterprise ID.

[0069] Furthermore, the enterprise data application management system starts enterprise data application management for the enterprise, and the acquisition module therein collects various public data related to the enterprise from various official websites, news websites, social platforms and other channels by calling API (Application Programming Interface), web crawlers and other technologies. In addition, according to the nature of the websites from which the various data come and the categories of the data themselves, the collected public data at least includes news information related to news reports, litigation information, social public opinion information, industry reports and policy documents, etc.

[0070] In addition, the enterprise data application management system also supports uploading internal data within the enterprise. After starting the enterprise data application management, the enterprise data application management system returns a prompt message to the enterprise, prompting the enterprise to upload internal data. If the enterprise has the need to upload internal data, it uploads the internal data based on the prompt message, and the acquisition module acquires the uploaded internal data. The internal data may at least include employee information, financial data and market data, so that the employee information reflects the educational level and years of work of the enterprise employees, and the financial data and market data reflect the financial status and market sales status of the enterprise.

[0071] The screening and verification module 20 is used to perform cleaning, deduplication and graphic conversion preprocessing on the public data, and perform default value verification on the internal data to obtain enterprise target data;

[0072] Understandably, various public data on the Internet may contain data that is actually irrelevant to the enterprise, as well as duplicate data, and may also contain data in the form of graphics. In this regard, preprocessing is performed through a screening and verification module, and the preprocessing at least includes cleaning, deduplication and image conversion. At the same time, for the internal data uploaded by the enterprise, one of the items may be missing. For such missing items, the internal data can be checked for default values ​​through the screening and verification module, and the default values ​​in the internal data can be determined through verification and completion operations can be performed. The internal data after the default value check and the public data after cleaning, deduplication and graphic conversion are then formed into the enterprise target data. In addition, in order to clean, dedupe and convert the public data, the screening and verification module is divided into a search unit, a screening unit and an identification unit: wherein,

[0073] Search unit 21: used for searching the public data for noise data irrelevant to the enterprise, and removing the noise data;

[0074] A screening unit 22, configured to determine duplicate data items in the public data, select any one of the duplicate data items as a reserved item, and delete other items of the duplicate data items except the reserved item;

[0075] The recognition unit 23 is used to recognize the graphics in the public data and determine the text information in the graphics.

[0076] Furthermore, the search unit searches for data in the public data that is not related to the enterprise as noise data, and removes the noise data from the public data. At the same time, the screening unit identifies duplicate data in the public data, selects any one item from each duplicate data and marks it as a retained item, and marks other data as deleted items, and then deletes the data marked as deleted items. For graphics in the public data, the recognition unit recognizes the text in the graphics to obtain the text information in the graphics. In this way, the preprocessing of the public data is achieved by removing noise data, deleting duplicate items, and identifying text information in the graphics.

[0077] An analysis module 30 is used to perform text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise;

[0078] Furthermore, a preset semantic network model is formed in advance through training in the enterprise data application management system, and the preset semantic network model can perform in-depth analysis on the text and identify the logical relationship and hidden meaning between the texts. The analysis module uses the preset semantic network model to perform in-depth text semantic analysis on the enterprise target data, and obtains basic information such as the enterprise scale and number of people, financial analysis results indicating the financial situation and development trend, market position analysis results indicating the position of the enterprise in the market, advantages and disadvantages information indicating the advantages and disadvantages of the enterprise, opportunity information of subsequent development opportunities, and risk information of possible risk points.

[0079] The generation module 40 is used to call a preset report template, and add the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template to generate and output the data application management report of the enterprise.

[0080] Furthermore, a template for generating a data application management report is pre-set in the enterprise data application management system, and the preset report template can be called through the generation module. Different areas in the template are filled with different modules, and basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are added to the preset report template and filled into the corresponding modules to generate the enterprise's data application management report. The report is then output to a terminal that is connected to the enterprise data application management system in a corresponding communication manner, such as a display device, so that the enterprise's managers can view it and better understand the enterprise's financial situation, market position, advantages, disadvantages, opportunities and risks.

[0081] Understandably, the relevant data of an enterprise is usually updated, such as news updates of major events of the enterprise, updates of relevant industry reports, updates of policies, etc. Such updates have an impact on the financial development, market position, advantages, disadvantages, opportunities, and risks of the enterprise. Therefore, in order to reflect the impact of data updates on the enterprise, this embodiment provides an update mechanism for the output data application management report of the enterprise. Specifically, the enterprise data application management system also includes:

[0082] The monitoring module a1 is used for monitoring whether there is first updated data corresponding to the public data and whether there is second updated data corresponding to the internal data at every preset period;

[0083] An updating module a2, configured to update the data application management report when the first update data and / or the second update data exist;

[0084] The display module a3 is used to control the display of the updated data application management report.

[0085] Furthermore, a preset period is pre-set, and at intervals of the preset period, the monitoring module monitors whether there is first update data corresponding to the public data, that is, determines whether there is news, policies, industry reports, etc. related to the enterprise within the preset period. At the same time, it monitors whether there is second update data corresponding to the internal data, that is, whether new internal data is uploaded. If there is first update data, or second update data, or both first update data and second update data exist, the update module updates the data application management report according to this type of update data. The update method can be similar to that of the data application management report, that is, the update data and the original data are pre-processed and the default value is checked, and the new enterprise target data is obtained by deep semantic analysis by the preset semantic network model, and new basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are obtained and filled into the preset report template to form a new data application management report, which is controlled by the display model to display and complete the update of the original data application management report.

[0086] It should be noted that multiple preset cycles will accumulate raw data for a long time, and there may be situations where the raw data is invalid or has no analytical significance. In this regard, a data collection time limit can be added during the preprocessing process. That is, the data collection time is determined during the preprocessing process. If the collection time exceeds the set time limit, the data is deleted to ensure the validity of the data.

[0087] Understandably, for the generated data application management report, personnel in the enterprise or personnel outside the enterprise may have the need to request the report to be sent for viewing. In order to ensure the security of transmission, this embodiment also provides a transmission encryption mechanism. Specifically, the enterprise data application management system also includes:

[0088] The first judgment module b1 is used to judge whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report when receiving the access request;

[0089] An encryption module b2, configured to generate an encryption key based on a preset algorithm when the account authority matches the report authority, and encrypt the data application management report based on the encryption key;

[0090] The transmission module b3 is used to transmit the encrypted data application management report to the access user account.

[0091] Furthermore, when the enterprise data application management system receives an access request for a data application management report, the judgment module obtains the access user account corresponding to the access request, and judges whether the account authority corresponding to the access user account matches the report authority corresponding to the data application management report. If they match, it means that the access user account can access the data application management report, and the data application management report can be sent to the access user account. In particular, for the security of the sending process, the sent data application management report needs to be encrypted.

[0092] Specifically, a preset algorithm is pre-set, and the encryption module generates an encryption key through the preset algorithm. The preset algorithm can be specifically referred to in the following formula (1).

[0093] X n =(K*X n-1 +N)mod W(1);

[0094] Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation.

[0095] Furthermore, the encryption module encrypts the data application management report by using an encryption key, and the transmission module transmits the encrypted data application management report to the user access account. The corresponding encryption and decryption key generation method is pre-set between the application software or webpage used to log in to the user access account and the enterprise data application management system. After the user access account receives the encrypted data application management report, it generates a decryption key by using the corresponding decryption key generation method, and decrypts the data application management report by using the decryption key to obtain the data application management report for viewing. In this way, the security of the data application management report transmission is ensured.

[0096] The enterprise data application management system implemented in this embodiment includes an acquisition module, a screening and verification module, an analysis module and a generation module: wherein the acquisition module is used to capture the public data of the enterprise and receive the internal data of the enterprise, wherein the public data at least includes news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data at least includes employee information, financial data and market data; the screening and verification module is used to pre-process the public data by cleaning, de-duplication and graphic conversion, and perform default value verification on the internal data to obtain the enterprise target data; the analysis module is used to perform text semantic analysis on the enterprise target data based on the preset semantic network model to obtain the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise; the generation module is used to call the preset report template, and add the obtained basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template, generate the enterprise data application management report and output it. In this way, various public data related to the enterprise are collected with the help of crawler technology, making the collected information more comprehensive. At the same time, the uploaded internal data is combined to generate enterprise target data. The preset semantic network model formed by a large amount of data training is used to perform in-depth text semantic analysis to obtain the company's basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information, which are added to the preset report template to generate the company's data application management report, so as to achieve comprehensive and accurate acquisition of information related to the enterprise and conduct efficient and accurate analysis.

[0097] For further information, please refer to Figure 2 Based on the first embodiment of the enterprise data application management system of the present invention, a second embodiment of the enterprise data application management system of the present invention is proposed.

[0098] The difference between the second embodiment of the enterprise data application management system and the first embodiment of the enterprise data application management system is that the enterprise data application management system further includes:

[0099] A collection module 50 is used to collect enterprise data samples, wherein the enterprise data samples include at least news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples;

[0100] A training module 60, used to train a preset network model according to the enterprise data sample and calculate a loss function value of the preset network model;

[0101] A second judgment module 70 is used to judge whether the loss function value is less than a preset loss threshold, and if it is less than the preset loss threshold, the preset network model is generated as the preset semantic network model;

[0102] An updating module 80, configured to update the weight values ​​and bias values ​​in the preset network model when the loss function value is not less than a preset loss threshold;

[0103] Execution module 90: for executing the step of training the preset network model according to the enterprise data sample for the updated preset network model.

[0104] This embodiment generates a preset semantic network model through training of a large number of enterprise data samples. Specifically, a large number of enterprise data samples are first collected through the collection module, and the enterprise data samples at least include news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples. Then, the preset network model is trained based on such enterprise data samples through the training module, and the loss function value of the preset network model is calculated after each training.

[0105] Furthermore, in order to reflect the loss size of the loss function value, a preset loss threshold is pre-set, and the loss function value is compared with the preset loss threshold to determine whether the loss function value is less than the preset loss threshold. If it is less than the preset loss threshold, it means that the error between the processing result obtained by the preset network model after training and processing the enterprise data sample and the reference result is small, and the preset network model has good language processing performance, so the preset network model is generated as a preset semantic network model. On the contrary, if the loss function value is determined to be greater than or equal to the preset loss threshold through comparison, it means that the language processing performance of the preset network model is poor. At this time, the weight value and bias value in the preset network model are updated through the update module, and the updated calculation formula can be seen in the following formula (2).

[0106]

[0107] Among them, w represents the updated weight value, w0 represents the weight value before updating, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before updating.

[0108] It should be noted that the preset network model can be set to include an input layer, a hidden layer and an output layer, and the corresponding weight value may include a first weight value between the hidden layer and the input layer, and a second weight value between the output layer and the hidden layer, and the bias value may include a first bias value between the hidden layer and the input layer, and a second bias value between the output layer and the hidden layer. The calculation formulas of the two weight values ​​and the two bias values ​​can be calculated by the above formula (2). When calculating the first weight value and the second weight value, w0 is the first weight value before updating and the second weight value before updating, respectively. When calculating the first bias value and the second bias value, b0 is the first bias value and the second bias value before updating, respectively.

[0109] Furthermore, after obtaining the updated weight value and bias value through calculation, the original weight value and bias value in the preset network model are replaced with the updated weight value and bias value to form an updated preset network model. Then, for the updated preset network model, the execution module executes the step of training the preset network model according to the enterprise data sample, and recalculates the loss function value to iteratively train the preset network model until the obtained loss function value is less than the preset loss threshold, stops the iterative training of the preset network model, and generates the preset network model into a preset semantic network model.

[0110] This embodiment generates a preset semantic network model through training of a large number of enterprise data samples, so that the preset semantic network model can accurately perform in-depth semantic analysis on enterprise target information, thereby improving the accuracy of obtaining basic enterprise information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information, thereby making the generated data application management report more accurate.

[0111] In addition, the present invention provides an enterprise data application management method, please refer to Figure 3 , Figure 3 It is a flowchart of the first embodiment of the enterprise data application management method of the present invention. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than here.

[0112] Specifically, the enterprise data application management method in this embodiment includes:

[0113] Step S10, capturing the public data of the enterprise and receiving the internal data of the enterprise, wherein the public data includes but is not limited to news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data includes but is not limited to employee information, financial data and market data;

[0114] Furthermore, when an enterprise needs to perform data classification application management, it initiates a management request to the enterprise data application management system of this embodiment. After receiving the management request, the enterprise data application management system reads the enterprise ID carried therein and determines the specific enterprise that needs data classification application management by the enterprise ID.

[0115] Furthermore, the enterprise data application management system starts enterprise data application management for the enterprise, and collects various public data related to the enterprise from various official websites, news websites, social platforms and other channels by calling API, web crawlers and other technologies. In addition, according to the nature of the websites from which the various data come and the categories of the data themselves, the collected public data at least includes news information related to news reports, litigation information, social public opinion information, industry reports and policy documents, etc.

[0116] In addition, the enterprise data application management system also supports uploading internal data within the enterprise. After starting the enterprise data application management, the enterprise data application management system returns a prompt message to the enterprise, prompting the enterprise to upload internal data. If the enterprise has the need to upload internal data, it uploads the internal data based on the prompt message, and the enterprise data application management system obtains the uploaded internal data. The internal data may at least include employee information, financial data and market data, so that the employee information reflects the education level and years of work of the enterprise employees, and the financial data and market data reflect the financial status and market sales status of the enterprise.

[0117] Step S20, pre-processing the public data by cleaning, de-duplication and graphic conversion, and performing default value verification on the internal data to obtain enterprise target data;

[0118] Understandably, the various public data on the Internet may contain data that is actually irrelevant to the enterprise, as well as duplicate data, and data in the form of graphics. At least pre-processing including cleaning, deduplication and image conversion is performed on this. At the same time, for the internal data uploaded by the enterprise, one of the items may be missing. For such missing items, it is necessary to perform a default value check on the internal data, determine the default value in the internal data through the check, and perform a completion operation. Then, the internal data after the default value check and the public data after cleaning, deduplication and graphic conversion are formed into the enterprise target data.

[0119] The steps of preprocessing the public data by cleaning, deduplication and graphic conversion include:

[0120] Step S21: searching for noise data irrelevant to the enterprise in the public data, and removing the noise data;

[0121] Step S22, for determining duplicate data items in the public data, selecting any one of the duplicate data items as a reserved item, and deleting the other items in the duplicate data items except the reserved item;

[0122] Step S23 is used to identify the graphics in the public data and determine the text information in the graphics.

[0123] Furthermore, data that is not related to the enterprise in the public data is searched as noise data, and the noise data is removed from the public data. At the same time, duplicate data in the public data is identified, and any one item from each duplicate data is selected as a retained item, and other data is identified as a deleted item, and then the data identified as the deleted item is deleted. For graphics in the public data, the text therein is identified to obtain the text information in the graphics. In this way, the preprocessing of public data is achieved by eliminating noise data, deleting duplicate items, and identifying text information in graphics.

[0124] Step S30, performing text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise;

[0125] Furthermore, a preset semantic network model is formed in advance through training in the enterprise data application management system, and the preset semantic network model can perform in-depth analysis on the text and identify the logical relationship and hidden meaning between the texts. Through the preset semantic network model, the enterprise target data is subjected to in-depth text semantic analysis to obtain basic information such as the enterprise scale and number of people, financial analysis results indicating the financial situation and development trend, market position analysis results indicating the position of the enterprise in the market, advantages and disadvantages information indicating the advantages and disadvantages of the enterprise, opportunity information of subsequent development opportunities, and risk information of possible risk points.

[0126] Step S40, calling a preset report template, and adding the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template, generating and outputting the data application management report of the enterprise.

[0127] Furthermore, a template for generating a data application management report is pre-set in the enterprise data application management system, and different areas in the template are filled with different modules. The preset report template is called, and basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are added to the preset report template and filled into the corresponding modules, so as to generate the enterprise's data application management report. The report is then output to a terminal that is connected to the enterprise data application management system in a corresponding communication manner, such as a display device, so that the enterprise's managers can view it and better understand the enterprise's financial situation, market position, advantages, disadvantages, opportunities and risks.

[0128] Understandably, the relevant data of an enterprise is usually updated, such as news updates of major events of the enterprise, updates of relevant industry reports, updates of policies, etc. Such updates have an impact on the financial development, market position, advantages, disadvantages, opportunities, and risks of the enterprise. Therefore, in order to reflect the impact of data updates on the enterprise, this embodiment provides an update mechanism for the output data application management report of the enterprise. Specifically, the step of generating and outputting the data application management report of the enterprise includes:

[0129] Step c1, monitoring whether there is first updated data corresponding to the public data and whether there is second updated data corresponding to the internal data at every preset period;

[0130] Step c2: if the first update data and / or the second update data exist, updating the data application management report;

[0131] Step c3, controlling the display of the updated data application management report.

[0132] Furthermore, a preset period is pre-set, and every time the preset period is spaced, the first update data corresponding to the public data is monitored, that is, whether news, policies, industry reports, etc. related to the enterprise appear within the preset period. At the same time, it is monitored whether the second update data corresponding to the internal data exists, that is, whether new internal data is uploaded. If the first update data exists, or the second update data exists, or the first update data and the second update data exist at the same time, the data application management report is updated according to this type of update data. The update method can be similar to that of the data application management report. That is, the update data and the original data are pre-processed and the default value is checked, and the new enterprise target data is obtained. The preset semantic network model is used for in-depth semantic analysis, and the new basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are obtained and filled into the preset report template to form a new data application management report for control display, and the update of the original data application management report is completed.

[0133] It should be noted that multiple preset cycles will accumulate raw data for a long time, and there may be situations where the raw data is invalid or has no analytical significance. In this regard, a data collection time limit can be added during the preprocessing process. That is, the data collection time is determined during the preprocessing process. If the collection time exceeds the set time limit, the data is deleted to ensure the validity of the data.

[0134] Understandably, for the generated data application management report, personnel in the enterprise or personnel outside the enterprise may have a request to send the report for viewing. In order to ensure the security of transmission, this embodiment is also provided with a transmission encryption mechanism. Specifically, after the step of generating and outputting the data application management report of the enterprise, it also includes:

[0135] Step d1, when receiving an access request, determining whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report;

[0136] Step d2: if the account authority matches the report authority, an encryption key is generated based on a preset algorithm, and the data application management report is encrypted based on the encryption key;

[0137] Step d3: transmitting the encrypted data application management report to the access user account.

[0138] Furthermore, when the enterprise data application management system receives an access request for a data application management report, it obtains the access user account corresponding to the access request and determines whether the account authority corresponding to the access user account matches the report authority corresponding to the data application management report. If they match, it means that the access user account can access the data application management report, and the data application management report can be sent to the access user account. In particular, for security during the sending process, the sent data application management report needs to be encrypted.

[0139] Specifically, a preset algorithm is pre-set, and the encryption key is generated by the preset algorithm. The preset algorithm can be specifically referred to in the following formula (1).

[0140] X n =(K*X n-1 +N)mod W(1);

[0141] Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation.

[0142] Furthermore, the data application management report is encrypted by an encryption key, and the encrypted data application management report is transmitted to the user access account. The corresponding encryption and decryption key generation method is pre-set between the application software or webpage used to log in to the user access account and the enterprise data application management system. After the user access account receives the encrypted data application management report, a decryption key is generated by a corresponding decryption key generation method, and the data application management report is decrypted by the decryption key to obtain the data application management report for viewing. In this way, the security of the data application management report transmission is ensured.

[0143] The enterprise data application management method implemented in this embodiment first captures the enterprise's public data and receives the enterprise's internal data. The public data at least includes news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data at least includes employee information, financial data and market data; then the public data is cleaned, deduplicated and pre-processed by graphic conversion, and the internal data is checked for default values ​​to obtain the enterprise's target data; then the enterprise target data is subjected to text semantic analysis based on a preset semantic network model to obtain the enterprise's basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information; the preset report template is called, and the obtained basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are added to the preset report template to generate and output the enterprise's data application management report. In this way, various public data related to the enterprise are collected with the help of crawler technology, making the collected information more comprehensive. At the same time, the uploaded internal data is combined to generate enterprise target data. The preset semantic network model formed by a large amount of data training is used to perform in-depth text semantic analysis to obtain the company's basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information, which are added to the preset report template to generate the company's data application management report, so as to achieve comprehensive and accurate acquisition of information related to the enterprise and conduct efficient and accurate analysis.

[0144] For further information, please refer to Figure 4 Based on the first embodiment of the enterprise data application management method of the present invention, a second embodiment of the enterprise data application management method of the present invention is proposed.

[0145] The difference between the second embodiment of the enterprise data application management method and the first embodiment of the enterprise data application management method is that the enterprise data application management system further includes:

[0146] Step S50, collecting enterprise data samples, wherein the enterprise data samples at least include news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples;

[0147] Step S60, training a preset network model according to the enterprise data sample, and calculating a loss function value of the preset network model;

[0148] Step S70, determining whether the loss function value is less than a preset loss threshold, and if so, generating the preset network model as the preset semantic network model;

[0149] Step S80, when the loss function value is not less than a preset loss threshold, updating the weight value and bias value in the preset network model;

[0150] Step S90: for the updated preset network model, executing the step of training the preset network model according to the enterprise data sample.

[0151] This embodiment generates a preset semantic network model by training a large number of enterprise data samples. Specifically, a large number of enterprise data samples are first collected, and the enterprise data samples at least include news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples. Then, the preset network model is trained based on such enterprise data samples, and the loss function value of the preset network model is calculated after each training.

[0152] Furthermore, in order to reflect the loss size of the loss function value, a preset loss threshold is pre-set, and the loss function value is compared with the preset loss threshold to determine whether the loss function value is less than the preset loss threshold. If it is less than the preset loss threshold, it means that the error between the processing result obtained by the preset network model after training and processing the enterprise data sample and the reference result is small, and the preset network model has good language processing performance, so the preset network model is generated as a preset semantic network model. On the contrary, if the loss function value is determined to be greater than or equal to the preset loss threshold through comparison, it means that the language processing performance of the preset network model is poor. At this time, the weight value and bias value in the preset network model are updated through the update module, and the updated calculation formula can be seen in the following formula (2).

[0153]

[0154] Among them, w represents the updated weight value, w0 represents the weight value before updating, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before updating.

[0155] It should be noted that the preset network model can be set to include an input layer, a hidden layer and an output layer, and the corresponding weight value may include a first weight value between the hidden layer and the input layer, and a second weight value between the output layer and the hidden layer, and the bias value may include a first bias value between the hidden layer and the input layer, and a second bias value between the output layer and the hidden layer. The calculation formulas of the two weight values ​​and the two bias values ​​can be calculated by the above formula (2). When calculating the first weight value and the second weight value, w0 is the first weight value before updating and the second weight value before updating, respectively. When calculating the first bias value and the second bias value, b0 is the first bias value and the second bias value before updating, respectively.

[0156] Furthermore, after obtaining the updated weight value and bias value through calculation, the original weight value and bias value in the preset network model are replaced with the updated weight value and bias value to form an updated preset network model. Then, for the updated preset network model, the step of training the preset network model according to the enterprise data sample is performed, and the loss function value is recalculated to iteratively train the preset network model until the obtained loss function value is less than the preset loss threshold, and the iterative training of the preset network model is stopped, and the preset network model is generated as a preset semantic network model.

[0157] This embodiment generates a preset semantic network model through training of a large number of enterprise data samples, so that the preset semantic network model can accurately perform in-depth semantic analysis on enterprise target information, thereby improving the accuracy of obtaining basic enterprise information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information, thereby making the generated data application management report more accurate.

[0158] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. An enterprise data application management system, characterized in that: The enterprise data application management system comprises: An acquisition module is used to capture the public data of an enterprise and receive the internal data of an enterprise. The public data includes but is not limited to news information, litigation information, social public opinion information, industry reports and policy documents, and the internal data includes but is not limited to employee information, financial data and market data; A screening and verification module is used to perform cleaning, deduplication and graphic conversion preprocessing on the public data, and to perform default value verification on the internal data to obtain enterprise target data; An analysis module, used to perform text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise; The generation module is used to call a preset report template, and add the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information to the preset report template to generate and output the data application management report of the enterprise.

2. The enterprise data application management system according to claim 1, characterized in that: The enterprise data application management system also includes: A monitoring module, used for monitoring whether there is first updated data corresponding to the public data and whether there is second updated data corresponding to the internal data at every preset period; an updating module, configured to update the data application management report when the first update data and / or the second update data exist; The display module is used to control the display of the updated data application management report.

3. The enterprise data application management system according to claim 1, characterized in that: The enterprise data application management system also includes: A first judgment module is used to, when receiving an access request, judge whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report; An encryption module is used to generate an encryption key based on a preset algorithm when the account authority matches the report authority, and encrypt the data application management report based on the encryption key, wherein the preset algorithm is: X n =(K*X n-1 +N)mod W; Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation; A transmission module is used to transmit the encrypted data application management report to the access user account.

4. The enterprise data application management system according to any one of claims 1 to 3, characterized in that: The enterprise data application management system also includes: A collection module, used to collect enterprise data samples, wherein the enterprise data samples include at least news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples; A training module, used to train a preset network model according to the enterprise data sample and calculate the loss function value of the preset network model; A second judgment module is used to judge whether the loss function value is less than a preset loss threshold, and if it is less than the preset loss threshold, the preset network model is generated as the preset semantic network model; An updating module is used to update the weight value and bias value in the preset network model when the loss function value is not less than the preset loss threshold, wherein the updating formula is: Wherein, w represents the updated weight value, w0 represents the weight value before the update, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before the update; Execution module: used for executing the step of training the preset network model according to the enterprise data sample for the updated preset network model.

5. The enterprise data application management system according to any one of claims 1 to 3, characterized in that: The screening and verification module comprises: A search unit: used for searching the public data for noise data irrelevant to the enterprise, and removing the noise data; A screening unit, used to determine duplicate data items in the public data, select any one of the duplicate data items as a reserved item, and delete other items of the duplicate data items except the reserved item; The recognition unit is used to recognize the graphics in the public data and determine the text information in the graphics.

6. An enterprise data application management method, characterized in that: The enterprise data application management method comprises: Capture the public data of enterprises and receive the internal data of enterprises. The public data includes but is not limited to news information, litigation information, public opinion information, industry reports and policy documents; the internal data includes but is not limited to employee information, financial data and market data; Pre-processing the public data by cleaning, de-duplication and graphic conversion, and performing default value verification on the internal data to obtain enterprise target data; Performing text semantic analysis on the enterprise target data based on a preset semantic network model to obtain basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information of the enterprise; The preset report template is called, and the basic information, financial analysis results, market position analysis results, advantage information, disadvantage information, opportunity information and risk information are added to the preset report template to generate and output the data application management report of the enterprise.

7. The enterprise data application management method according to claim 6, characterized in that: The step of generating and outputting the data application management report of the enterprise includes: At every preset period, monitoring whether there is first updated data corresponding to the public data, and whether there is second updated data corresponding to the internal data; If the first update data and / or the second update data exist, updating the data application management report; Controlling display of the updated data application management report.

8. The enterprise data application management method according to claim 6, characterized in that: The step of generating and outputting the data application management report of the enterprise further includes: When receiving an access request, determining whether the account authority of the access user account corresponding to the access request matches the report authority corresponding to the data application management report; If the account authority matches the report authority, an encryption key is generated based on a preset algorithm, and the data application management report is encrypted based on the encryption key, wherein the preset algorithm is: X n =(K*X n-1 +N)mod W; Among them, X n represents the currently generated encryption key, K represents the preset multiplier parameter, X n-1 Indicates the encryption key generated by the last encryption, N indicates the preset increment parameter, W indicates the preset modulus parameter, and mod indicates the modulus operation; The encrypted data application management report is transmitted to the access user account.

9. The enterprise data application management method according to any one of claims 6 to 8, characterized in that: The step of performing text semantic analysis on the enterprise target data based on the preset semantic network model includes: Collecting enterprise data samples, wherein the enterprise data samples include at least news samples, litigation samples, public opinion samples, industry report samples, employee information samples, financial data samples and market data samples; Training a preset network model according to the enterprise data sample, and calculating a loss function value of the preset network model; Determine whether the loss function value is less than a preset loss threshold, and if so, generate the preset network model as the preset semantic network model; If the loss function value is not less than the preset loss threshold, the weight value and bias value in the preset network model are updated, wherein the update formula is: Wherein, w represents the updated weight value, w0 represents the weight value before the update, τ represents the learning rate of the preset network model, Loss represents the loss function value, b represents the updated bias value, and b0 represents the bias value before the update; For the updated preset network model, a step of training the preset network model according to the enterprise data sample is performed.

10. The enterprise data application management method according to any one of claims 6 to 8, characterized in that: The steps of pre-processing the public data by cleaning, deduplication and graphic conversion include: Searching for noise data in the public data that is irrelevant to the enterprise data, and removing the noise data; Determine duplicate data items in the public data, select any one of the duplicate data items as a reserved item, and delete other items of the duplicate data items except the reserved item; The graphics in the public data are identified to determine the text information in the graphics.

Images