Network public opinion management system and method based on deep learning

Abstract

The present invention discloses a network public opinion management system and method based on deep learning, which relates to the field of network management technology, and sets a network information chain, wherein the network information chain includes an information pool and a guide bar; the network information chain and the network platform form a network closed loop; multiple network information are obtained by uploading through multiple network terminals, multiple network information are classified by type through a differentiation pool to obtain multiple type information, and the type information is bound to the corresponding network information to obtain the first information; the first information is processed to obtain the first uploaded information, and the first uploaded information is transmitted to the management pool through the guide bar for screening and management. The present invention can ensure data combing and quality management during the data upload process on the network platform, ensure the quality of the network information of the network platform, and maintain the network quality environment.

Description

Technical Field

[0001] The present invention relates to the technical field of network management, and in particular to a network public opinion management system and method based on deep learning. Background Art

[0002] Internet public opinion management refers to a management activity that monitors, analyzes, evaluates and responds to public opinion on a specific event, topic or issue on the Internet; the control of Internet information is also an important part of social stability management. At present, it is difficult to ensure the early processing of information in the management of Internet information. There will be a situation where the information is directly input into the network platform without being processed in the early stage, causing public opinion pressure, and then subsequent rescue can be carried out, resulting in a certain amount of low-quality information flooding on the Internet. Summary of the invention

[0003] The purpose of the present invention is to provide a network public opinion management system and method based on deep learning to solve the shortcomings of the background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solution: a network public opinion management method based on deep learning, comprising the following steps:

[0005] Setting a network information chain, wherein the network information chain includes an information pool and a guide bar;

[0006] Form a network closed loop between the network information chain and the network platform;

[0007] A plurality of network information is obtained by uploading through a plurality of network terminals, the plurality of network information is classified by type through a differentiation pool to obtain a plurality of type information, and the type information is bound with the corresponding network information to obtain the first information;

[0008] Processing the first information to obtain first uploaded information, and transmitting the first uploaded information to the management pool through the guide bar for screening and management;

[0009] Based on the management pool, the corresponding first uploaded information is put into the corresponding preview pool to obtain the preview result, and the network information is disseminated on the network platform according to the preview result.

[0010] In a preferred embodiment, the step of setting up a network information link includes:

[0011] Set up differentiation pool, management pool and rehearsal pool as information pools;

[0012] The differentiation pool, the management pool and the rehearsal pool are linked by guide bars;

[0013] Use the information pool and guide bar as the network information chain.

[0014] In a preferred embodiment, the step of associating the differentiation pool, the management pool and the preview pool through the guide bar comprises:

[0015] Setting a communication channel between the differentiation pool, the management pool and the rehearsal pool in the order of the differentiation pool, the management pool and the rehearsal pool;

[0016] The corresponding differentiation pool, management pool and preview pool are respectively marked as feature information, wherein the feature information includes a coding and random coding library;

[0017] A plurality of guide nodes are arranged in the communication channel, and the plurality of guide nodes are connected for communication;

[0018] The communication channel after the communication connection between the bootstrap nodes is used as the bootstrap bar.

[0019] In a preferred embodiment, the step of forming a network closed loop with the network information chain and the network platform includes:

[0020] Establish a communication channel between the preview pool and the network platform;

[0021] Establish communication channels between the network end and the network platform and differentiation pool respectively;

[0022] A network closed loop is formed based on the communication connection between the network terminal, network platform and network information chain.

[0023] In a preferred embodiment, the steps of obtaining multiple network information by uploading through multiple network terminals, classifying the multiple network information by type through a differentiation pool to obtain multiple type information, and binding the type information with the corresponding network information to obtain the first information include:

[0024] Based on big data, historical network information and corresponding type information are obtained as historical data, and a convolutional neural network is trained based on the historical data to obtain a learning model;

[0025] A plurality of type windows are set in the differentiation pool corresponding to the type information, and the type windows are bound to the type information;

[0026] Upload multiple network information to the differentiation pool based on multiple network terminals, and classify the multiple network information based on the learning model in the differentiation pool to obtain multiple types of information;

[0027] The type information is bound with the corresponding network information as the first information.

[0028] In a preferred embodiment, the step of processing the first information to obtain the first uploaded information, and transmitting the first uploaded information to the management pool through the guide bar for screening and management includes:

[0029] Acquire information points in the first information, wherein the information points are multiple keywords in the first information;

[0030] A plurality of extraction points are set corresponding to the first information, a code corresponding to the differentiation pool is assigned to the first information and the extraction points are randomly assigned, a random code is assigned based on a random code library, the random code is associated with a plurality of keywords in the first information and the extraction points are assigned, and the first uploaded information is obtained;

[0031] storing the codes, random codes and information points corresponding to the random codes in the guide points between the differentiation pool and the management pool;

[0032] Establishing the traction information of the code, random code and information point corresponding to the random code in the first uploaded information and the guide point respectively, wherein the traction information is the extraction point position of the guide point pair code, random code and information point corresponding to the random code;

[0033] The first uploaded information is transmitted through the guide bar based on the type information, and the guide node extracts the code, the random code and the information point corresponding to the random code based on the traction information;

[0034] The extracted codes, random codes and information points corresponding to the random codes are compared with the codes, random codes and information points corresponding to the random codes stored in the guide points to obtain comparison results. When the comparison results are consistent, the first uploaded information is allowed to be transmitted through the guide node corresponding to the guide bar, and based on the type information, it is transmitted to the corresponding management pool, and the first uploaded information is sensitively screened and managed in the management pool.

[0035] In a preferred embodiment, the step of putting the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain a preview result, and disseminating the network information on the network platform according to the preview result includes:

[0036] The number of management pools and the number of rehearsal pools are in a one-to-one relationship;

[0037] Assigning a code corresponding to the management pool to the first uploaded information in the management pool, generating a random code based on a random code library in the management pool, and associating the random code with an information point corresponding to the random code;

[0038] storing the coding, random coding and information points corresponding to the random coding in the guide points between the management pool and the preview pool;

[0039] The code, random code and information point corresponding to the random code in the management pool are replaced with the code, random code and information point corresponding to the random code in the differentiation pool at the corresponding extraction point to obtain the second uploaded information;

[0040] The second uploaded information is transmitted through the guide bar, and the guide node extracts the code, the random code and the information point corresponding to the random code in the second uploaded information;

[0041] Compare the extracted code, random code and information point corresponding to the random code with the code, random code and information point corresponding to the random code stored in the guide point to obtain a comparison result, and when the comparison results are consistent, allow the second upload information to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the corresponding preview pool;

[0042] Previewing the propagation of the second uploaded information on the network platform in the preview pool to obtain a preview result, wherein the preview result includes the propagation efficiency and the load pressure of the network platform on the propagation efficiency;

[0043] The second upload information corresponding to the preview result that meets the preset conditions is uploaded to the network platform as security information.

[0044] The present invention also provides a network public opinion management system based on deep learning, including:

[0045] A setting module, used for setting a network information chain, wherein the network information chain includes an information pool and a guide bar;

[0046] The construction module is connected with the setting module and is used to form a network closed loop between the network information chain and the network platform;

[0047] A first uploading module, connected to the building module, is used to obtain multiple network information by uploading through multiple network terminals, classify the multiple network information by type through the differentiation pool to obtain multiple type information, and bind the type information with the corresponding network information to obtain the first information;

[0048] A second uploading module, connected to the first uploading module, is used to process the first information to obtain first uploaded information, and transmit the first uploaded information to the management pool through the guide bar for screening and management;

[0049] The propagation module is connected to the second upload module and is used to put the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain the preview result, and propagate the network information on the network platform according to the preview result.

[0050] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0051] The present invention transmits the first upload information through the guide bar based on the type information during the transmission process. The guide node extracts the information points corresponding to the coding, random coding and random coding based on the traction information. There are multiple management pools here, and one management pool corresponds to one type information, which are independent of each other. The extracted coding, random coding and random coding corresponding information points are compared with the coding, random coding and random coding corresponding information points stored in the guide point to obtain a comparison result. When the comparison results are consistent, the first upload information is allowed to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the management pool. This can ensure data combing and quality management during the data upload process on the network platform, ensure the quality of network information on the network platform, and maintain the network quality environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0053] Figure 1 The figure is a flow chart of the method of the present invention.

[0054] Figure 2 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] Example 1, please refer to Figure 1 As shown, the network public opinion management method based on deep learning described in this embodiment includes the following steps:

[0057] S1. Setting a network information chain, wherein the network information chain includes an information pool and a guide bar;

[0058] S2. Form a network closed loop between the network information chain and the network platform;

[0059] S3, obtaining multiple network information by uploading through multiple network terminals, classifying the multiple network information by type through a differentiation pool to obtain multiple type information, and binding the type information with the corresponding network information to obtain first information;

[0060] S4, processing the first information to obtain first uploaded information, and transmitting the first uploaded information to the management pool through the guide bar for screening and management;

[0061] S5. Based on the management pool, the corresponding first uploaded information is put into the corresponding preview pool to obtain a preview result, and the network information is disseminated on the network platform according to the preview result;

[0062] As described in the above steps S1-S5, with the development of society, the control of network information is also an important part of social stability management. At present, it is difficult to ensure the pre-processing of information in the management of network information. There will be a situation where the information is directly put into the network platform without pre-processing, causing public opinion pressure, and then subsequent rescue can be carried out, resulting in a certain amount of low-quality network information flooding. In this application, the first uploaded information is transmitted through the guide bar based on the type information during the transmission process, and the guide node extracts the information points corresponding to the coding, random coding and random coding based on the traction information. There are multiple management pools here, and one management pool corresponds to one type of information, which are independent of each other. The extracted coding, random coding and random coding corresponding information points are compared with the coding, random coding and random coding corresponding information points stored in the guide point to obtain a comparison result. When the comparison results are consistent, the first uploaded information is allowed to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the management pool, which can ensure a data combing and quality management during the data upload process on the network platform, ensure the quality of network information on the network platform, and maintain the network quality environment;

[0063] In one embodiment, the step S1 of setting the network information link includes:

[0064] S11. Set up a differentiation pool, a management pool and a rehearsal pool as information pools;

[0065] S12, associating the differentiation pool, the management pool and the preview pool through a guide bar;

[0066] S13, using the information pool and the guide bar as a network information chain;

[0067] As described in the above steps S11-S13, there are multiple ports on the network data publishing platform for publishing network information and viewing network information. Among them, in order to manage network public opinion in the release of information, it is first necessary to prepare for the release of network information. Here, a differentiation pool, a management pool and a preview pool are set as information pools, and the network information is processed in the early stage through the information pool. After that, the information is operated through the information pool and then published on the network platform. The differentiation pool, the management pool and the preview pool are associated through a guide bar, and the differentiation pool, the management pool and the preview pool are associated through a guide bar, which has a good information processing effect;

[0068] In one embodiment, the step S12 of associating the differentiation pool, the management pool and the preview pool through the guide bar comprises:

[0069] S121, setting a communication channel between the differentiation pool, the management pool and the preview pool in the order of the differentiation pool, the management pool and the preview pool;

[0070] S122, marking the corresponding differentiation pool, management pool and preview pool respectively as feature information, wherein the feature information includes a code and a random code library;

[0071] S123, setting a plurality of guide nodes in the communication channel, and establishing communication connection between the plurality of guide nodes;

[0072] S124, using the communication channel after the communication connection between the guiding nodes as the guiding bar;

[0073] As described in the above steps S121-S124, when a process is performed on the network information before it is released, it needs to be transmitted in sequence in the order of the differentiation pool, the management pool and the rehearsal pool. Therefore, it is necessary to set a communication channel in the order of the differentiation pool, the management pool and the rehearsal pool. The communication channel is set between the differentiation pool, the management pool and the rehearsal pool, and then the characteristic information is set corresponding to the differentiation pool, the management pool and the rehearsal pool, wherein the characteristic information is the code set corresponding to the differentiation pool, the management pool and the rehearsal pool, and at the same time, a random coding library is set in the differentiation pool, the management pool and the rehearsal pool, and the random coding library is used for subsequent coding association of the network information to facilitate whether the subsequently transmitted network information is processed through the network information chain. The network information can be processed through the differentiation pool, the management pool and the rehearsal pool, and then the code of the marking information pool is marked corresponding to the network information, and the network information is given a random code and a random code library is generated based on the random code, which can be used for subsequent verification in the process of network information transmission;

[0074] In one embodiment, the step S2 of forming a network closed loop with the network information chain and the network platform includes:

[0075] S21. Establish a communication channel between the preview pool and the network platform;

[0076] S22, establishing communication channels between the network end and the network platform and the differentiation pool respectively;

[0077] S23, forming a network closed loop based on the communication connection between the network terminal, the network platform and the network information chain;

[0078] As described in the above steps S21-S23, after the network information chain is set up, the differentiation pool in the information pool is connected to the network end, the rehearsal pool is connected to the network platform, and then the network platform is connected to the network end. The network platform is a one-way communication connection that provides network information to the network end. Finally, a network closed loop is formed according to the communication connection between the network end, the network platform and the network information chain. The network platform is a platform for the network end to obtain network information and publish network information.

[0079] In one embodiment, the step S3 of obtaining multiple network information by uploading through multiple network terminals, classifying the multiple network information by type through a differentiation pool to obtain multiple type information, and binding the type information with the corresponding network information to obtain the first information includes:

[0080] S31, acquiring historical network information and corresponding type information as historical data based on the big data, and training a convolutional neural network based on the historical data to obtain a learning model;

[0081] S32, setting a plurality of type windows in the differentiation pool corresponding to the type information, and binding the type windows with the type information;

[0082] S33, uploading multiple network information to the differentiation pool based on multiple network terminals, and classifying the multiple network information based on the learning model in the differentiation pool to obtain multiple types of information;

[0083] S34, binding the type information with the corresponding network information as the first information;

[0084] As described in the above steps S31-S34, the differentiation pool here is used to classify network information by type. Here, first, the type information existing in the existing society is obtained based on big data, and then the historical network information and the corresponding type information are obtained as historical data. The convolutional neural network is trained based on the historical data to obtain a learning model. In the differentiation pool, the network information uploaded by the network end can be classified through the learning model to obtain type information. Based on the type information of the big data, a corresponding type window is set in the differentiation pool, which is used to transmit network information of the same type of information to the management pool as a transmission port. Before transmission, the type information is bound to the corresponding network information for transmission, which can be used for subsequent transmission accuracy and security of network data, ensuring that the data has been processed by the network information chain, avoiding other network information attacks. The network information chain is not directly uploaded to the network platform through processing, which has a good network data management function and a subsequent public opinion management effect;

[0085] In one embodiment, the step S4 of processing the first information to obtain the first uploaded information and transmitting the first uploaded information to the management pool through the guide bar for screening and management includes:

[0086] S41, obtaining information points in the first information, where the information points are multiple keywords in the first information;

[0087] S42, setting a plurality of extraction points corresponding to the first information, assigning the code corresponding to the differentiation pool to the first information and randomly assigning the extraction points, assigning random codes based on the random code library, associating the random codes with a plurality of keywords in the first information and assigning the extraction points, and obtaining the first uploaded information;

[0088] S43, storing the code, random code and information point corresponding to the random code in the guide point between the differentiation pool and the management pool;

[0089] S44, establishing traction information of the code, random code and information point corresponding to the random code in the first uploaded information and the guide point, wherein the traction information is the extraction point position of the guide point pair code, random code and information point corresponding to the random code;

[0090] S45, transmitting the first uploaded information through the guide bar based on the type information, and the guide node extracting the code, the random code and the information point corresponding to the random code based on the traction information;

[0091] S46, comparing the extracted code, random code and information point corresponding to the random code with the code, random code and information point corresponding to the random code stored in the guide point to obtain a comparison result, and when the comparison results are consistent, allowing the first uploaded information to be transmitted through the guide node corresponding to the guide bar, and transmitting it to the corresponding management pool based on the type information, and performing sensitive screening management on the first uploaded information in the management pool;

[0092] As described in the above steps S41-S46, the network information is first classified in the differentiation pool to obtain the corresponding type information. Before transmitting through the guide bar, it is necessary to extract keywords from the content of the first information as information points, set multiple extraction points corresponding to the first information, assign the code corresponding to the differentiation pool to the first information and randomly assign extraction points, assign random codes based on the random code library, associate the random codes with multiple keywords in the first information and assign extraction points to obtain the first uploaded information, where the extraction points are the record codes, random codes and multiple keywords in the first information corresponding to the random codes, the coding is to correspond to the first information after being processed by the differentiation pool, establish the traction information of the code, random code and information points corresponding to the random code in the first uploaded information, and the guide point, wherein the traction information is the extraction point of the guide point pair code, random code and information points corresponding to the random code, and the random code and information points corresponding to the random code are the information binding corresponding to the content of the first information, here through the random code The association between the information points can help us understand whether the keyword information has been tampered with, and ensure the upload security of the data network. During the transmission process, the first uploaded information is transmitted through the guide bar based on the type information. The guide node extracts the information points corresponding to the code, random code and random code based on the traction information. There are multiple management pools here, and one management pool corresponds to one type of information, which are independent of each other. The extracted information points corresponding to the code, random code and random code are compared with the information points corresponding to the code, random code and random code stored in the guide point to obtain the comparison results. When the comparison results are consistent, the first uploaded information is allowed to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the management pool for sensitive screening management (for some social industry confidential information or a series of information with greater harm, it can be used as a shielding condition to shield and cannot be uploaded to the network platform). This can ensure data combing and quality management during the upload process of the network platform, ensure the quality of network information on the network platform, and maintain the network quality environment.

[0093] In one embodiment, the step S5 of putting the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain the preview result, and propagating the network information on the network platform according to the preview result includes:

[0094] S51. The number of management pools and the number of preview pools are in a one-to-one relationship;

[0095] S52, assigning a code corresponding to the management pool to the first uploaded information in the management pool, generating a random code based on a random code library in the management pool, and associating the random code with an information point corresponding to the random code;

[0096] S53, storing the code, random code and information points corresponding to the random code in the guide point between the management pool and the preview pool;

[0097] S54, replacing the codes, random codes and information points corresponding to the random codes in the management pool with the codes, random codes and information points corresponding to the random codes in the differentiation pool at corresponding extraction points to obtain second uploaded information;

[0098] S55, transmitting the second uploaded information through the guide bar, and the guide node extracting the code, the random code and the information point corresponding to the random code in the second uploaded information;

[0099] S56, comparing the extracted code, random code and information point corresponding to the random code with the code, random code and information point corresponding to the random code stored in the guide point to obtain a comparison result, and when the comparison results are consistent, allowing the second upload information to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the corresponding preview pool;

[0100] S57, performing a preview on the propagation of the second uploaded information on the network platform in the preview pool to obtain a preview result, wherein the preview result includes a propagation efficiency and a load pressure of the network platform on the propagation efficiency;

[0101] S58, uploading the second uploaded information corresponding to the preview result that meets the preset conditions to the network platform as security information;

[0102] As described in the above steps S51-S58, the operation of transmitting the network information in the differentiation pool to the management pool is the same as that of transmitting the network information in the management pool to the rehearsal pool. The only difference is the coding and random coding, which can ensure that the uploaded network information is processed through the network information chain, and avoid some information directly skipping the information processing and directly uploading it to the network platform. On the one hand, it is impossible to count the types of information and it is difficult to carry out subsequent corresponding management. On the other hand, it is impossible to manage the public opinion of the data and it is difficult to ensure the quality of the uploaded information.

[0103] Example 2, please refer to Figure 2 As shown, the network public opinion management system based on deep learning described in this embodiment includes:

[0104] A setting module, used for setting a network information chain, wherein the network information chain includes an information pool and a guide bar;

[0105] The construction module is connected with the setting module and is used to form a network closed loop between the network information chain and the network platform;

[0106] A first uploading module, connected to the building module, is used to obtain multiple network information by uploading through multiple network terminals, classify the multiple network information by type through the differentiation pool to obtain multiple type information, and bind the type information with the corresponding network information to obtain the first information;

[0107] A second uploading module, connected to the first uploading module, is used to process the first information to obtain first uploaded information, and transmit the first uploaded information to the management pool through the guide bar for screening and management;

[0108] A propagation module, connected to the second upload module, is used to put the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain a preview result, and propagate the network information on the network platform according to the preview result;

[0109] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A network public opinion management method based on deep learning, characterized in that: The following steps are involved: Setting a network information chain, wherein the network information chain includes an information pool and a guide bar; The step of setting the network information chain includes: Set up differentiation pool, management pool and rehearsal pool as information pools; The differentiation pool, the management pool and the rehearsal pool are linked by guide bars; The step of associating the differentiation pool, the management pool and the preview pool through the guide bar comprises: Setting a communication channel between the differentiation pool, the management pool and the rehearsal pool in the order of the differentiation pool, the management pool and the rehearsal pool; The corresponding differentiation pool, management pool and preview pool are respectively marked as feature information, wherein the feature information includes a coding and random coding library; A plurality of guide nodes are arranged in the communication channel, and the plurality of guide nodes are connected for communication; The communication channel after the communication connection between the guiding nodes is used as the guiding bar; Use the information pool and guide bar as a network information chain; Form a network closed loop between the network information chain and the network platform; A plurality of network information is obtained by uploading through a plurality of network terminals, the plurality of network information is classified by type through a differentiation pool to obtain a plurality of type information, and the type information is bound with the corresponding network information to obtain the first information; Processing the first information to obtain first uploaded information, and transmitting the first uploaded information to the management pool through the guide bar for screening and management; The step of processing the first information to obtain the first uploaded information, and transmitting the first uploaded information to the management pool through the guide bar for screening and management, includes: Acquire information points in the first information, wherein the information points are multiple keywords in the first information; A plurality of extraction points are set corresponding to the first information, a code corresponding to the differentiation pool is assigned to the first information and the extraction points are randomly assigned, a random code is assigned based on a random code library, the random code is associated with a plurality of keywords in the first information and the extraction points are assigned, and the first uploaded information is obtained; storing the codes, random codes and information points corresponding to the random codes in the guide points between the differentiation pool and the management pool; Establishing the traction information of the code, random code and information point corresponding to the random code in the first uploaded information and the guide point respectively, wherein the traction information is the extraction point position of the guide point pair code, random code and information point corresponding to the random code; The first uploaded information is transmitted through the guide bar based on the type information, and the guide node extracts the code, the random code and the information point corresponding to the random code based on the traction information; Compare the extracted codes, random codes and information points corresponding to the random codes with the codes, random codes and information points corresponding to the random codes stored in the guide points to obtain a comparison result. When the comparison results are consistent, allow the first uploaded information to be transmitted through the guide node corresponding to the guide bar, and transmit it to the corresponding management pool based on the type information, and perform sensitive screening management on the first uploaded information in the management pool; Based on the management pool, the corresponding first uploaded information is put into the corresponding preview pool to obtain a preview result, and the network information is disseminated on the network platform according to the preview result; The step of putting the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain a preview result, and disseminating the network information on the network platform according to the preview result includes: The number of management pools and the number of rehearsal pools are in a one-to-one relationship; Assigning a code corresponding to the management pool to the first uploaded information in the management pool, generating a random code based on a random code library in the management pool, and associating the random code with an information point corresponding to the random code; storing the coding, random coding and information points corresponding to the random coding in the guide points between the management pool and the preview pool; The code, random code and information point corresponding to the random code in the management pool are replaced with the code, random code and information point corresponding to the random code in the differentiation pool at the corresponding extraction point to obtain the second uploaded information; The second uploaded information is transmitted through the guide bar, and the guide node extracts the code, the random code and the information point corresponding to the random code in the second uploaded information; Compare the extracted code, random code and information point corresponding to the random code with the code, random code and information point corresponding to the random code stored in the guide point to obtain a comparison result, and when the comparison results are consistent, allow the second upload information to be transmitted through the guide node corresponding to the guide bar until it is transmitted to the corresponding preview pool; Previewing the propagation of the second uploaded information on the network platform in the preview pool to obtain a preview result, wherein the preview result includes the propagation efficiency and the load pressure of the network platform on the propagation efficiency; The second upload information corresponding to the preview result that meets the preset conditions is uploaded to the network platform as security information.

2. The network public opinion management method based on deep learning according to claim 1 is characterized by: The step of forming a network closed loop with the network information chain and the network platform includes: Establish a communication channel between the preview pool and the network platform; Establish communication channels between the network end and the network platform and differentiation pool respectively; A network closed loop is formed based on the communication connection between the network terminal, network platform and network information chain.

3. The network public opinion management method based on deep learning according to claim 1 is characterized by: The steps of obtaining multiple network information by uploading through multiple network terminals, classifying the multiple network information by type through a differentiation pool to obtain multiple type information, and binding the type information with the corresponding network information to obtain the first information include: Based on big data, historical network information and corresponding type information are obtained as historical data, and a convolutional neural network is trained based on the historical data to obtain a learning model; A plurality of type windows are set in the differentiation pool corresponding to the type information, and the type windows are bound to the type information; Upload multiple network information to the differentiation pool based on multiple network terminals, and classify the multiple network information based on the learning model in the differentiation pool to obtain multiple types of information; The type information is bound with the corresponding network information as the first information.

4. A network public opinion management system based on deep learning, used to implement the network public opinion management method based on deep learning according to any one of claims 1 to 3, characterized in that: include: A setting module, used for setting a network information chain, wherein the network information chain includes an information pool and a guide bar; The construction module is connected with the setting module and is used to form a network closed loop between the network information chain and the network platform; A first uploading module, connected to the building module, is used to obtain multiple network information by uploading through multiple network terminals, classify the multiple network information by type through the differentiation pool to obtain multiple type information, and bind the type information with the corresponding network information to obtain the first information; A second uploading module, connected to the first uploading module, is used to process the first information to obtain first uploaded information, and transmit the first uploaded information to the management pool through the guide bar for screening and management; The propagation module is connected to the second upload module and is used to put the corresponding first uploaded information into the corresponding preview pool based on the management pool to obtain the preview result, and propagate the network information on the network platform according to the preview result.

Images