A method for monitoring and processing network public opinion data based on cluster analysis

Abstract

The present invention relates to the field of network public opinion data monitoring and analysis, and in particular to a network public opinion data monitoring and processing method based on cluster analysis, comprising: using real-time network public opinion data to establish multi-level data features of network public opinion data; using the multi-level data features of the network public opinion data to obtain basic classification results of the real-time network public opinion data based on cluster analysis; using the basic classification results of the real-time network public opinion data to obtain network public opinion data monitoring and processing results, based on cluster analysis, the historical semantic features and network communication features of the real-time data are output as independent comparison data respectively, firstly avoiding the misjudgment of the results of single semantics or single source IP, and secondly improving the complete process of implementing the plan through iterative optimization process, improving the monitoring efficiency, and gradually expanding the cluster analysis database according to the implementation time of the plan, improving the generalization ability of the plan implementation and the accuracy of processing large quantities of data.

Description

Technical Field

[0001] The present invention relates to the field of network public opinion data monitoring and analysis, and in particular to a network public opinion data monitoring and processing method based on cluster analysis. Background Art

[0002] Compared with traditional public opinion, online public opinion is more timely, interactive and extensive, and can quickly aggregate the views of a large number of netizens and form a strong social influence. At the same time, online public opinion is also easily affected by false information and emotional expressions, and needs to be guided and supervised to ensure its healthy development. As an important factor affecting public opinion and decision-making, dynamic monitoring of online public opinion is particularly critical. Since the Internet is full of various opinions, views and comments, these information converge into a powerful torrent of public opinion. Manual collection, screening and organization of this information alone is obviously unable to cope with the huge amount of information and the rapid update speed. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention provides a method for monitoring and processing network public opinion data based on cluster analysis. By independently establishing analysis and comparison processes for the semantic features and network features of real-time network public opinion data, problems such as errors caused by outputting analysis results with a single feature can be avoided.

[0004] To achieve the above object, the present invention provides a method for monitoring and processing network public opinion data based on cluster analysis, comprising:

[0005] S1. Use real-time online public opinion data to establish multi-level data features of online public opinion data;

[0006] S2. Obtain basic classification results of real-time network public opinion data based on cluster analysis using the multi-level data features of the network public opinion data;

[0007] S3. Obtain the network public opinion data monitoring and processing results using the basic classification results of the real-time network public opinion data.

[0008] Preferably, the method of using real-time network public opinion data to establish multi-level data features of network public opinion data includes:

[0009] Collect real-time online public opinion data;

[0010] Using the real-time online public opinion data to establish network data features of the online public opinion data;

[0011] Using the real-time network public opinion data to establish public opinion semantic features of the network public opinion data;

[0012] The network data features and public opinion semantic features of the network public opinion data are used as multi-level data features of the network public opinion data.

[0013] Furthermore, the network data features of the network public opinion data established by using the real-time network public opinion data include:

[0014] According to the real-time network public opinion data, the corresponding server interface address and the user local interface address are respectively obtained;

[0015] Acquire the corresponding server interface data volume according to the server interface address;

[0016] Acquire the corresponding user local interface data volume according to the user local interface address;

[0017] Acquire the corresponding secondary network communication line according to the server interface address and the user local interface address;

[0018] Acquire the corresponding communication protocol according to the current network communication line;

[0019] The server interface address, user local interface address, server interface data volume, user local interface data volume, current network communication line and communication protocol are used as network data features of network public opinion data.

[0020] Furthermore, using the real-time online public opinion data to establish public opinion semantic features of the online public opinion data includes:

[0021] Using the real-time network public opinion data to perform word segmentation processing to obtain a text vocabulary set of the real-time network public opinion data;

[0022] Using the text vocabulary set of the real-time network public opinion data to perform text cleaning processing to obtain a collection of text keywords of the real-time network public opinion data;

[0023] Using the text keyword collection to establish a text key word semantic sequence based on real-time network public opinion data;

[0024] Using the text keyword collection, respectively obtaining part-of-speech tagging sequences corresponding to the text keyword collection;

[0025] The text keyword collection, text key word semantic sequence and part-of-speech tagging sequence of the real-time network public opinion data are used as the public opinion semantic features of the network public opinion data.

[0026] Furthermore, the basic classification results of the real-time network public opinion data are obtained based on cluster analysis using the multi-level data features of the network public opinion data, including:

[0027] S2-1, using the multi-level data features of the network public opinion data to correspond to the network data features based on cluster analysis to obtain the basic network feature classification results of the real-time network public opinion data;

[0028] S2-2, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features based on cluster analysis to obtain the basic public opinion semantic feature classification results of the real-time network public opinion data;

[0029] S2-3. Use the basic network feature classification results and the basic public opinion semantic feature classification results of the real-time network public opinion data as the basic classification results of the real-time network public opinion data.

[0030] Furthermore, the basic network feature classification results of the real-time network public opinion data are obtained by using the multi-level data features of the network public opinion data to correspond to the network data features based on cluster analysis, including:

[0031] S2-1-1, obtaining corresponding historical normal network data features and historical abnormal network data features respectively according to the multi-level data features of the network public opinion data corresponding to the network data features;

[0032] S2-1-2, using the historical normal network data features as the initial network data features to check the verification tag;

[0033] S2-1-3, using the server interface address of the historical abnormal network data feature as the abnormal network data feature address label;

[0034] S2-1-4, using the user local interface addresses of the historical abnormal network data characteristics to establish abnormal user local interface address regional clusters of the historical abnormal network data characteristics based on cluster analysis according to the user local interface address corresponding regions;

[0035] S2-1-5, using the server interface data volume of the historical abnormal network data feature to establish an abnormal network data feature capacity label according to the corresponding time;

[0036] S2-1-6, using the user local interface data volume of the historical abnormal network data characteristics to establish an abnormal user local interface data volume time cluster of the historical abnormal network data characteristics based on cluster analysis according to the corresponding time;

[0037] S2-1-7, using the current network communication line and communication protocol of the historical abnormal network data characteristics as abnormal network data characteristics review and verification labels;

[0038] S2-1-8. Use the initial network data feature verification label, abnormal network data feature address label, abnormal user local interface address area cluster, abnormal network data feature capacity label, abnormal user local interface data volume time cluster and abnormal network data feature review verification label as the basic network feature classification result of real-time network public opinion data.

[0039] Furthermore, the basic public opinion semantic feature classification results of the real-time network public opinion data are obtained by using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features based on cluster analysis, including:

[0040] S2-2-1, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features, respectively obtain the corresponding historical normal public opinion semantic features and historical abnormal public opinion semantic features;

[0041] S2-2-2, using the normal public opinion semantic features as the initial public opinion semantic features to check and verify the label;

[0042] S2-2-3, using the text keyword collection of the historical abnormal public opinion semantic features to obtain first text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection;

[0043] S2-2-4, using the text keyword collection of the historical normal public opinion semantic features to obtain second text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection;

[0044] S2-2-5, determine whether the first text keyword cluster is consistent with the second text keyword cluster, if so, remove the second text keyword cluster from the first text keyword cluster to obtain the abnormal public opinion semantic feature vocabulary label, otherwise, use the first text keyword cluster as the abnormal public opinion semantic feature vocabulary label;

[0045] S2-2-6. Using the semantic sequence of key words in the text of the historical abnormal public opinion semantic feature, a first public opinion semantic feature moment label is established based on cluster analysis according to the corresponding moment;

[0046] S2-2-7, using the text key word semantic sequence of the historical normal public opinion semantic feature to establish a second public opinion semantic feature moment label based on cluster analysis according to the corresponding moment;

[0047] S2-2-8. Determine whether the first public opinion semantic feature moment label and the second public opinion semantic feature moment label have the same moment. If so, remove the second public opinion semantic feature moment label from the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label. Otherwise, use the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label.

[0048] S2-2-9. Use the initial public opinion semantic feature verification label, abnormal public opinion semantic feature vocabulary label and abnormal public opinion semantic feature moment label as the basic public opinion semantic feature classification result of real-time network public opinion data.

[0049] Furthermore, the network public opinion data monitoring and processing results obtained by using the basic classification results of the real-time network public opinion data include:

[0050] S3-1, using the basic classification results of the real-time network public opinion data to perform initial verification processing to obtain an initial verification processing result of the basic classification result;

[0051] S3-2, performing verification processing on the initial verification processing result of the basic classification result to obtain the network public opinion data monitoring processing result.

[0052] Further, the initial verification processing is performed using the basic classification results of the real-time network public opinion data to obtain the initial verification processing results of the basic classification results, including:

[0053] S3-1-1, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature address label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-2, otherwise, directly execute S3-1-5;

[0054] S3-1-2, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface address area cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-3, otherwise, directly execute S3-1-5;

[0055] S3-1-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface data volume time cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-4, otherwise, directly execute S3-1-5;

[0056] S3-1-4, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature review and verification label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, the initial verification processing result of the basic classification result is abnormal, otherwise, execute S3-1-5;

[0057] S3-1-5. Determine whether the network data features of the real-time network public opinion data correspond to the initial network data feature verification labels of the basic network feature classification results corresponding to the basic classification results of the real-time network public opinion data. If so, the initial verification processing result of the basic classification result is normal. Otherwise, use the network data features of the real-time network public opinion data to update the historical normal network data features and the historical abnormal network data features respectively, and return to S2-1-1.

[0058] Furthermore, the network public opinion data monitoring processing results obtained by checking and verifying the initial verification processing results of the basic classification results include:

[0059] S3-2-1, determine whether the initial verification processing result of the basic classification result is normal, if so, execute S3-2-2, otherwise, directly execute S3-2-3;

[0060] S3-2-2, judging whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification label of the basic public opinion semantic feature classification result, if so, the network public opinion data monitoring and processing result is normal, otherwise, executing S3-2-3;

[0061] S3-2-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal public opinion semantic feature vocabulary label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-2-4, otherwise, directly execute S3-2-6;

[0062] S3-2-4, determine whether the public opinion semantic features of the real-time network public opinion data correspond to the abnormal public opinion semantic feature moment label of the basic public opinion semantic feature classification result of the basic public opinion semantic feature classification result of the real-time network public opinion data, if so, the network public opinion data monitoring and processing result is abnormal, otherwise, execute S3-2-6;

[0063] S3-2-6. Determine whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification label of the basic public opinion semantic feature classification result corresponding to the basic classification result of the real-time network public opinion data. If so, the network public opinion data monitoring and processing result is normal. Otherwise, use the public opinion semantic features of the real-time network public opinion data to update the historical normal public opinion semantic features and the historical abnormal public opinion semantic features respectively, and return to S2-2-1.

[0064] Compared with the closest prior art, the present invention has the following beneficial effects:

[0065] Based on cluster analysis, the historical semantic features and network communication features of real-time data are output as independent comparison data respectively. This first avoids the misjudgment of results of single semantics or single source IP. Secondly, through iterative optimization process, the complete process of implementation of the plan is improved, the monitoring efficiency is improved, and the cluster analysis database is gradually expanded according to the implementation time of the plan, which improves the generalization ability of the plan implementation and the accuracy of processing large amounts of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a flow chart of a method for monitoring and processing network public opinion data based on cluster analysis provided by the present invention. DETAILED DESCRIPTION

[0067] The specific implementation modes of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0068] 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.

[0069] Embodiment 1:

[0070] The present invention provides a method for monitoring and processing network public opinion data based on cluster analysis. Figure 1 As shown, including:

[0071] S1. Use real-time online public opinion data to establish multi-level data features of online public opinion data;

[0072] S2. Obtain basic classification results of real-time network public opinion data based on cluster analysis using the multi-level data features of the network public opinion data;

[0073] S3. Obtain the network public opinion data monitoring and processing results using the basic classification results of the real-time network public opinion data.

[0074] S1 specifically includes:

[0075] S1-1. Collect real-time online public opinion data;

[0076] S1-2, using the real-time network public opinion data to establish network data features of the network public opinion data;

[0077] S1-3, using the real-time network public opinion data to establish public opinion semantic features of the network public opinion data;

[0078] S1-4. Utilize the network data features and public opinion semantic features of the network public opinion data as multi-level data features of the network public opinion data.

[0079] S1-2 specifically includes:

[0080] S1-2-1. Obtain the corresponding server interface address and user local interface address respectively according to the real-time network public opinion data;

[0081] S1-2-2. Obtain the corresponding server interface data volume according to the server interface address;

[0082] S1-2-3, obtaining the corresponding user local interface data volume according to the user local interface address;

[0083] S1-2-4, obtaining the corresponding secondary network communication line according to the server interface address and the user local interface address;

[0084] S1-2-5, obtaining a corresponding communication protocol according to the current network communication line;

[0085] S1-2-6. Use the server interface address, user local interface address, server interface data volume, user local interface data volume, current network communication line and communication protocol as network data features of network public opinion data.

[0086] S1-3 specifically includes:

[0087] S1-3-1, using the real-time network public opinion data to perform word segmentation processing to obtain a text vocabulary set of the real-time network public opinion data;

[0088] S1-3-2, using the text vocabulary set of the real-time network public opinion data to perform text cleaning processing to obtain a collection of text keywords of the real-time network public opinion data;

[0089] S1-3-3, using the text keyword collection to establish a text key word semantic sequence based on real-time network public opinion data;

[0090] S1-3-4, using the text keyword collection to respectively obtain part-of-speech tagging sequences corresponding to the text keyword collection;

[0091] S1-3-5. Utilize the text keyword collection, text key word semantic sequence and part-of-speech tagging sequence of the real-time network public opinion data as the public opinion semantic features of the network public opinion data.

[0092] S2 specifically includes:

[0093] S2-1, using the multi-level data features of the network public opinion data to correspond to the network data features based on cluster analysis to obtain the basic network feature classification results of the real-time network public opinion data;

[0094] S2-2, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features based on cluster analysis to obtain the basic public opinion semantic feature classification results of the real-time network public opinion data;

[0095] S2-3. Use the basic network feature classification results and the basic public opinion semantic feature classification results of the real-time network public opinion data as the basic classification results of the real-time network public opinion data.

[0096] S2-1 specifically includes:

[0097] S2-1-1, obtaining corresponding historical normal network data features and historical abnormal network data features respectively according to the multi-level data features of the network public opinion data corresponding to the network data features;

[0098] S2-1-2, using the historical normal network data features as initial network data features to check and verify the tag;

[0099] S2-1-3, using the server interface address of the historical abnormal network data feature as the abnormal network data feature address label;

[0100] S2-1-4, using the user local interface addresses of the historical abnormal network data characteristics to establish abnormal user local interface address regional clusters of the historical abnormal network data characteristics based on cluster analysis according to the user local interface address corresponding regions;

[0101] S2-1-5, using the server interface data volume of the historical abnormal network data feature to establish an abnormal network data feature capacity label according to the corresponding time;

[0102] S2-1-6, using the user local interface data volume of the historical abnormal network data characteristics to establish an abnormal user local interface data volume time cluster of the historical abnormal network data characteristics based on cluster analysis according to the corresponding time;

[0103] S2-1-7, using the current network communication line and communication protocol of the historical abnormal network data characteristics as abnormal network data characteristics review and verification labels;

[0104] S2-1-8. Use the initial network data feature verification label, abnormal network data feature address label, abnormal user local interface address area cluster, abnormal network data feature capacity label, abnormal user local interface data volume time cluster and abnormal network data feature review verification label as the basic network feature classification result of real-time network public opinion data.

[0105] In the present embodiment, a method for monitoring and processing network public opinion data based on cluster analysis is used to perform analysis according to the corresponding region using the user local interface address. In the specific implementation, it can be defined according to the overall scope corresponding to the data. When the data volume as a whole corresponds to the province, the scope of the corresponding region should be all municipal areas within the province, or when the data volume as a whole corresponds to the municipal scope, the region is defined as the district-level region. Through the secondary administrative division of the region corresponding to the data volume, batch and accurate processing of cluster analysis can be achieved.

[0106] S2-2 specifically includes:

[0107] S2-2-1, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features, respectively obtain the corresponding historical normal public opinion semantic features and historical abnormal public opinion semantic features;

[0108] S2-2-2, using the normal public opinion semantic features as the initial public opinion semantic features to check and verify the label;

[0109] S2-2-3, using the text keyword collection of the historical abnormal public opinion semantic features to obtain first text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection;

[0110] S2-2-4, using the text keyword collection of the historical normal public opinion semantic features to obtain second text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection;

[0111] S2-2-5, determine whether the first text keyword cluster is consistent with the second text keyword cluster, if so, remove the second text keyword cluster from the first text keyword cluster to obtain the abnormal public opinion semantic feature vocabulary label, otherwise, use the first text keyword cluster as the abnormal public opinion semantic feature vocabulary label;

[0112] S2-2-6. Using the semantic sequence of key words in the text of the historical abnormal public opinion semantic feature, a first public opinion semantic feature moment label is established based on cluster analysis according to the corresponding moment;

[0113] S2-2-7, using the text key word semantic sequence of the historical normal public opinion semantic feature to establish a second public opinion semantic feature moment label based on cluster analysis according to the corresponding moment;

[0114] S2-2-8. Determine whether the first public opinion semantic feature moment label and the second public opinion semantic feature moment label have the same moment. If so, remove the second public opinion semantic feature moment label from the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label. Otherwise, use the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label.

[0115] S2-2-9. Use the initial public opinion semantic feature verification label, abnormal public opinion semantic feature vocabulary label and abnormal public opinion semantic feature moment label as the basic public opinion semantic feature classification result of real-time network public opinion data.

[0116] S3 specifically includes:

[0117] S3-1, using the basic classification results of the real-time network public opinion data to perform initial verification processing to obtain an initial verification processing result of the basic classification result;

[0118] S3-2, performing verification processing on the initial verification processing result of the basic classification result to obtain the network public opinion data monitoring processing result.

[0119] S3-1 specifically includes:

[0120] S3-1-1, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature address label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-2, otherwise, directly execute S3-1-5;

[0121] S3-1-2, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface address area cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-3, otherwise, directly execute S3-1-5;

[0122] S3-1-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface data volume time cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-4, otherwise, directly execute S3-1-5;

[0123] S3-1-4, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature review and verification label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, the initial verification processing result of the basic classification result is abnormal, otherwise, execute S3-1-5;

[0124] S3-1-5. Determine whether the network data features of the real-time network public opinion data correspond to the initial network data feature verification labels of the basic network feature classification results corresponding to the basic classification results of the real-time network public opinion data. If so, the initial verification processing result of the basic classification result is normal. Otherwise, use the network data features of the real-time network public opinion data to update the historical normal network data features and the historical abnormal network data features respectively, and return to S2-1-1.

[0125] S3-2 specifically includes:

[0126] S3-2-1, determine whether the initial verification processing result of the basic classification result is normal, if so, execute S3-2-2, otherwise, directly execute S3-2-3;

[0127] S3-2-2, judging whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification label of the basic public opinion semantic feature classification result, if so, the network public opinion data monitoring and processing result is normal, otherwise, executing S3-2-3;

[0128] S3-2-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal public opinion semantic feature vocabulary label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-2-4, otherwise, directly execute S3-2-6;

[0129] S3-2-4, determine whether the public opinion semantic features of the real-time network public opinion data correspond to the abnormal public opinion semantic feature moment label of the basic public opinion semantic feature classification result of the basic public opinion semantic feature classification result of the real-time network public opinion data, if so, the network public opinion data monitoring and processing result is abnormal, otherwise, execute S3-2-6;

[0130] S3-2-6. Determine whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification label of the basic public opinion semantic feature classification result corresponding to the basic classification result of the real-time network public opinion data. If so, the network public opinion data monitoring and processing result is normal. Otherwise, use the public opinion semantic features of the real-time network public opinion data to update the historical normal public opinion semantic features and the historical abnormal public opinion semantic features respectively, and return to S2-2-1.

[0131] In this embodiment, a method for monitoring and processing network public opinion data based on cluster analysis, in the final output of monitoring and processing results, is to compare the corresponding data of each level of features of real-time network public opinion data with the corresponding data of pre-established basic network data feature classification results and basic public opinion semantic feature classification results one by one. Only after the complete comparison is passed, it is confirmed that the monitoring and processing of the public opinion data is normal.

[0132] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for monitoring and processing network public opinion data based on cluster analysis, characterized in that: include: S1. Use real-time online public opinion data to establish multi-level data features of online public opinion data; S1-1. Collect real-time online public opinion data; S1-2, using the real-time network public opinion data to establish network data features of the network public opinion data; S1-2-1. Obtain the corresponding server interface address and user local interface address respectively according to the real-time network public opinion data; S1-2-2. Obtain the corresponding server interface data volume according to the server interface address; S1-2-3, obtaining the corresponding user local interface data volume according to the user local interface address; S1-2-4, obtaining the corresponding secondary network communication line according to the server interface address and the user local interface address; S1-2-5, obtaining a corresponding communication protocol according to the current network communication line; S1-2-6, using the server interface address, user local interface address, server interface data volume, user local interface data volume, current network communication line and communication protocol as network data features of network public opinion data; S1-3, using the real-time network public opinion data to establish public opinion semantic features of the network public opinion data; S1-4, using the network data features and public opinion semantic features of the network public opinion data as multi-level data features of the network public opinion data; S2. Obtain basic classification results of real-time network public opinion data based on cluster analysis using the multi-level data features of the network public opinion data; S2-1, using the multi-level data features of the network public opinion data to correspond to the network data features based on cluster analysis to obtain the basic network feature classification results of the real-time network public opinion data; S2-1-1, obtaining corresponding historical normal network data features and historical abnormal network data features respectively according to the multi-level data features of the network public opinion data corresponding to the network data features; S2-1-2, using the historical normal network data features as initial network data features to check and verify the tag; S2-1-3, using the server interface address of the historical abnormal network data feature as the abnormal network data feature address label; S2-1-4, using the user local interface addresses of the historical abnormal network data characteristics to establish abnormal user local interface address regional clusters of the historical abnormal network data characteristics based on cluster analysis according to the user local interface address corresponding regions; S2-1-5, using the server interface data volume of the historical abnormal network data feature to establish an abnormal network data feature capacity label according to the corresponding time; S2-1-6, using the user local interface data volume of the historical abnormal network data characteristics to establish an abnormal user local interface data volume time cluster of the historical abnormal network data characteristics based on cluster analysis according to the corresponding time; S2-1-7, using the current network communication line and communication protocol of the historical abnormal network data characteristics as abnormal network data characteristics review and verification labels; S2-1-8, using the initial network data feature verification label, abnormal network data feature address label, abnormal user local interface address area cluster, abnormal network data feature capacity label, abnormal user local interface data volume time cluster and abnormal network data feature review verification label as the basic network feature classification result of real-time network public opinion data; S2-2, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features based on cluster analysis to obtain the basic public opinion semantic feature classification results of the real-time network public opinion data; S2-2-1, using the multi-level data features of the network public opinion data to correspond to the public opinion semantic features, respectively obtain the corresponding historical normal public opinion semantic features and historical abnormal public opinion semantic features; S2-2-2, using the historical normal public opinion semantic features as the initial public opinion semantic features to check and verify the label; S2-2-3, using the text keyword collection of the historical abnormal public opinion semantic features to obtain first text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection; S2-2-4, using the text keyword collection of the historical normal public opinion semantic features to obtain second text keyword collection clusters corresponding to each subset of the text keyword collection based on cluster analysis according to each subset corresponding to the text keyword collection; S2-2-5, determine whether the first text keyword cluster is consistent with the second text keyword cluster, if so, remove the second text keyword cluster from the first text keyword cluster to obtain the abnormal public opinion semantic feature vocabulary label, otherwise, use the first text keyword cluster as the abnormal public opinion semantic feature vocabulary label; S2-2-6. Using the semantic sequence of key words in the text of the historical abnormal public opinion semantic feature, a first public opinion semantic feature moment label is established based on cluster analysis according to the corresponding moment; S2-2-7, using the text key word semantic sequence of the historical normal public opinion semantic feature to establish a second public opinion semantic feature moment label based on cluster analysis according to the corresponding moment; S2-2-8. Determine whether the first public opinion semantic feature moment label and the second public opinion semantic feature moment label have the same moment. If so, remove the second public opinion semantic feature moment label from the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label. Otherwise, use the first public opinion semantic feature moment label as the abnormal public opinion semantic feature moment label. S2-2-9, using the initial public opinion semantic feature verification label, abnormal public opinion semantic feature vocabulary label and abnormal public opinion semantic feature moment label as the basic public opinion semantic feature classification result of the real-time network public opinion data; S2-3, using the basic network feature classification results and the basic public opinion semantic feature classification results of the real-time network public opinion data as the basic classification results of the real-time network public opinion data; S3. Obtain the network public opinion data monitoring and processing results using the basic classification results of the real-time network public opinion data.

2. The method for monitoring and processing network public opinion data based on cluster analysis according to claim 1, characterized in that: Using the real-time online public opinion data to establish public opinion semantic features of the online public opinion data includes: Using the real-time network public opinion data to perform word segmentation processing to obtain a text vocabulary set of the real-time network public opinion data; Using the text vocabulary set of the real-time network public opinion data to perform text cleaning processing to obtain a collection of text keywords of the real-time network public opinion data; Using the text keyword collection to establish a text key word semantic sequence based on real-time network public opinion data; Using the text keyword collection, respectively obtaining part-of-speech tagging sequences corresponding to the text keyword collection; The text keyword collection, text key word semantic sequence and part-of-speech tagging sequence of the real-time network public opinion data are used as the public opinion semantic features of the network public opinion data.

3. The method for monitoring and processing network public opinion data based on cluster analysis according to claim 2, characterized in that: The network public opinion data monitoring and processing results obtained by using the basic classification results of the real-time network public opinion data include: S3-1, using the basic classification results of the real-time network public opinion data to perform initial verification processing to obtain an initial verification processing result of the basic classification result; S3-2, performing verification processing on the initial verification processing result of the basic classification result to obtain the network public opinion data monitoring processing result.

4. The method for monitoring and processing network public opinion data based on cluster analysis according to claim 3, characterized in that: The initial verification processing results of the basic classification results obtained by performing initial verification processing on the basic classification results of the real-time network public opinion data include: S3-1-1, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature address label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-2, otherwise, directly execute S3-1-5; S3-1-2, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface address area cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-3, otherwise, directly execute S3-1-5; S3-1-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal user local interface data volume time cluster of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-1-4, otherwise, directly execute S3-1-5; S3-1-4, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal network data feature review and verification label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, the initial verification processing result of the basic classification result is abnormal, otherwise, execute S3-1-5; S3-1-5. Determine whether the network data features of the real-time network public opinion data correspond to the initial network data feature verification labels of the basic network feature classification results corresponding to the basic classification results of the real-time network public opinion data. If so, the initial verification processing result of the basic classification result is normal. Otherwise, use the network data features of the real-time network public opinion data to update the historical normal network data features and the historical abnormal network data features respectively, and return to S2-1-1.

5. The method for monitoring and processing network public opinion data based on cluster analysis according to claim 4, characterized in that: The network public opinion data monitoring processing results obtained by checking and verifying the initial verification processing results of the basic classification results include: S3-2-1, determine whether the initial verification processing result of the basic classification result is normal, if so, execute S3-2-2, otherwise, directly execute S3-2-3; S3-2-2, judging whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification label of the basic public opinion semantic feature classification result, if so, the network public opinion data monitoring and processing result is normal, otherwise, executing S3-2-3; S3-2-3, determine whether the network data feature of the real-time network public opinion data corresponds to the abnormal public opinion semantic feature vocabulary label of the basic network feature classification result of the basic classification result of the real-time network public opinion data, if so, execute S3-2-4, otherwise, directly execute S3-2-6; S3-2-4, determine whether the public opinion semantic features of the real-time network public opinion data correspond to the abnormal public opinion semantic feature moment label of the basic public opinion semantic feature classification result of the basic public opinion semantic feature classification result of the real-time network public opinion data, if so, the network public opinion data monitoring and processing result is abnormal, otherwise, execute S3-2-6; S3-2-6. Determine whether the public opinion semantic features of the real-time network public opinion data correspond to the initial public opinion semantic feature verification labels of the basic public opinion semantic feature classification results corresponding to the basic classification results of the real-time network public opinion data. If so, the network public opinion data monitoring and processing results are normal. Otherwise, use the public opinion semantic features of the real-time network public opinion data to update the historical normal public opinion semantic features and the historical abnormal public opinion semantic features respectively, and return to S2-2-1.

Images