Network public opinion data change analysis method based on time sequence analysis

Abstract

The invention relates to the field of network public opinion data change analysis, in particular to a time sequence analysis-based network public opinion data change analysis method, which comprises the following steps of: acquiring network public opinion data characteristics by utilizing real-time network public opinion data; according to the real-time network public opinion data and the network public opinion data characteristics, establishing a network public opinion time sequence data chain based on time sequence analysis; and obtaining a network public opinion data change analysis result according to the network public opinion time sequence data chain, introducing time sequence analysis into change process processing, establishing an analysis basis by extracting multiple features of the network public opinion data and various hardware addresses associated with the features, and subsequently performing multi-moment multi-feature cross judgment and screening, so that the network public opinion data change analysis result is obtained. The change condition of the network public opinion data is obtained through step-by-step analysis, the change part is selected in a targeted mode, the public opinion data change screening efficiency is improved, and it is ensured that the final output result is stable and reliable.

Description

Technical Field

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

[0002] Nowadays, in the Internet era, various social software emerges in an endless stream, which is a medium that reflects online public opinion. Nowadays, the public opinions published in various social media are mostly false, malicious and negative. In order to create a better network environment, conventional means are used to monitor real-time public opinion data through data crawling. However, for this subdivided monitoring field, monitoring and monitoring by a single means alone cannot obtain effective monitoring results in the current situation of alternative words and complex implicit meanings. At the same time, the reasonable screening of the behavior monitoring of continuous speech and evaluation of specific IPs and network accounts has a certain impact on the judgment of subsequent public opinion data. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention provides a method for analyzing changes in network public opinion data based on time series analysis. By effectively processing real-time network public opinion data, a method for collaboratively analyzing changes in public opinion data with multiple features at multiple times based on the principle of time series analysis is proposed, thereby effectively improving the analysis efficiency and the accuracy of the final output results.

[0004] To achieve the above object, the present invention provides a method for analyzing changes in network public opinion data based on time series analysis, comprising: S1. Use real-time online public opinion data to obtain online public opinion data characteristics; S2. Establishing a network public opinion time series data chain based on time series analysis according to the real-time network public opinion data and network public opinion data characteristics; S3. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain.

[0005] Preferably, the method of obtaining network public opinion data features using real-time network public opinion data includes: Obtain the corresponding time of real-time network public opinion data as the analysis benchmark time t; Obtain text data and user IP address of the real-time network public opinion data; Obtain corresponding page views and platform IP addresses according to the real-time network public opinion data; The analysis reference time t, text data of real-time network public opinion data and user IP address are used as network public opinion data features.

[0006] Furthermore, establishing a network public opinion time series data chain based on time series analysis according to the real-time network public opinion data and network public opinion data features includes: S2-1, obtaining text data features of real-time network public opinion data according to the text data corresponding to the network public opinion data features; S2-2, using the real-time network public opinion data and text data features to establish a basic network public opinion time series data chain according to the analysis reference time t of the network public opinion data features; S2-3. Utilize the basic network public opinion time series data chain to perform verification processing based on time series analysis according to the user IP address of the network public opinion data characteristics to obtain the network public opinion time series data chain.

[0007] Further, obtaining text data features of real-time network public opinion data according to the network public opinion data features corresponding to the text data includes: Using the network public opinion data features to perform text cleaning on the corresponding text data to obtain source text data; Using the source text data to perform word segmentation processing to obtain source text word segmentation data; Using the source text word segmentation data to obtain the word vector of the source text word segmentation data; Establishing a word vector weight of the source text word segmentation data according to the word vector of the source text word segmentation data; Using the word vectors and word vector weights of the source text word segmentation data as text data features of real-time network public opinion data; The text cleaning process is to remove noise data in the text.

[0008] Furthermore, using the real-time network public opinion data and text data features to establish a basic network public opinion time series data chain according to the analysis reference time t of the network public opinion data features includes: Using the real-time network public opinion data, according to the analysis reference time t of the network public opinion data characteristics, a real-time network public opinion data chain at the analysis reference time t, time t+1 and time t+2 is established; Using the text data features, a text data feature data chain at the analysis reference time t, time t+1 and time t+2 is established according to the analysis reference time t of the network public opinion data features; Using the real-time network public opinion data link and the text data feature data link to establish a data-text feature time mapping according to the corresponding time; According to the data-text feature time mapping between the analysis reference time t and the time t+1, the similarity of the word vectors of the source text word segmentation data is obtained as the first data link auxiliary feature; According to the data-text feature time mapping at time t+1 and time t+2, the similarity of the word vectors of the source text word segmentation data is obtained as the second data link auxiliary feature; The real-time network public opinion data chain, the text data feature data chain, the first data chain auxiliary feature and the second data chain auxiliary feature are used as the basic network public opinion time series data chain.

[0009] Further, the basic network public opinion time series data chain is used to perform verification processing based on the user IP address of the network public opinion data characteristics based on time series analysis to obtain the network public opinion time series data chain, including: S2-3-1. Based on the user IP address of the network public opinion data characteristics, the user hardware information, the current communication protocol information and the server interface address of the network public opinion data characteristics are obtained based on the time series analysis; S2-3-2, determine whether the user hardware information corresponds to the user IP address, if so, obtain the communication line between the user hardware information and the user IP address, and execute S2-3-3, otherwise, return to S2-3-1; S2-3-3, determine whether the current communication protocol information corresponds to the communication line, if so, obtain the process data corresponding to the communication line and execute S2-3-4, otherwise, return to S2-3-1; S2-3-4, determine whether the process data corresponds to the server interface address, if so, execute S2-3-5, otherwise, return to S2-3-1; S2-3-5, determine whether the number of executions of S2-3-1 at the current moment is greater than 4, if so, abandon the processing, otherwise, retain the current basic network public opinion time series data chain as the network public opinion time series data chain; The communication protocol information at that time is the communication protocol between the user IP address and the server interface address.

[0010] Furthermore, the network public opinion data change analysis results obtained according to the network public opinion time series data chain include: S3-1, using the network public opinion time series data chain to establish the characteristics of network public opinion data change trends; S3-2. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain and the network public opinion data change trend characteristics.

[0011] Furthermore, using the network public opinion time series data chain to establish the network public opinion data change trend characteristics includes: Establish a first data change direction by using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at the analysis reference time t and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1; Establish a second data change direction by using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1 and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+2; Establish a first text feature change direction by using the text data feature data chain corresponding to the network public opinion time series data chain at the analysis reference time t and the text data feature data chain corresponding to the network public opinion time series data chain at time t+1; Establish a second text feature change direction by using the text data feature data chain corresponding to the network public opinion time series data chain at time t+1 and the text data feature data chain corresponding to the network public opinion time series data chain at time t+2; Using the network public opinion time series data link to correspond to the first data link auxiliary feature and the second data link auxiliary feature, establish a data link auxiliary feature spacing change direction; The first data change direction, the second data change direction, the first text feature change direction, the second text feature change direction and the data link auxiliary feature spacing change direction are used as network public opinion data change trend features.

[0012] Furthermore, the network public opinion data change analysis results obtained based on the network public opinion time series data chain and the network public opinion data change trend characteristics include: S3-2-1. Use the word vector weight corresponding to the network public opinion time series data chain to establish a data label value; S3-2-2, determine whether the pointing directions of the first data change point and the second data change point are consistent, if so, execute S3-2-3, otherwise, respectively obtain the pointing directions of the data tag value corresponding to the first data change point and the pointing directions of the data tag value corresponding to the second data change point, and execute S3-2-4; S3-2-3, determine whether the average value of the data tag value corresponding to the first data change is consistent with the average value of the data tag value corresponding to the second data change. If so, output the network public opinion time series data chain as the network public opinion data change analysis result. Otherwise, execute S3-2-4; S3-2-4, determine whether the direction of the data tag value corresponding to the first data change is consistent with the direction of the first data change. If so, execute S3-2-5; otherwise, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result, and execute S3-2-5; S3-2-5. Determine whether the direction of the data tag value corresponding to the second data change is consistent with the direction of the second data change. If so, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result. Otherwise, output the real-time network public opinion data corresponding to the analysis base time t, time t+1 and time t+2 as the network public opinion data change analysis result. Among them, the pointing direction includes rising, falling and maintaining, the rising means that the data at the adjacent next moment is greater than the data at the adjacent previous moment, the falling means that the data at the adjacent next moment is less than the data at the adjacent previous moment, and the maintaining means that the data at the adjacent next moment is equal to the data at the adjacent previous moment.

[0013] Compared with the closest prior art, the present invention has the following beneficial effects: Introduce time series analysis into the change process processing, and establish an analysis basis by extracting various features of online public opinion data and various hardware addresses associated with them. Then, conduct cross-judgment and screening of multiple features at multiple times in the future, gradually analyze and obtain the changes in online public opinion data, and select the changed parts in a targeted manner to improve the efficiency of screening changes in public opinion data and ensure the stability and reliability of the final output results. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a flow chart of a method for analyzing changes in network public opinion data based on time series analysis provided by the present invention. DETAILED DESCRIPTION

[0015] The specific implementation modes of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0016] 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.

[0017] Embodiment 1:

[0018] The present invention provides a method for analyzing network public opinion data changes based on time series analysis. Figure 1 As shown, including: S1. Use real-time online public opinion data to obtain online public opinion data characteristics; S2. Establishing a network public opinion time series data chain based on time series analysis according to the real-time network public opinion data and network public opinion data characteristics; S3. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain.

[0019] S1 specifically includes: S1-1, obtain the corresponding time of real-time network public opinion data as the analysis reference time t; S1-2, obtaining the text data and user IP address of the real-time network public opinion data; S1-3. Obtain corresponding page views and platform IP address according to the real-time network public opinion data; S1-4, using the analysis reference time t, the text data of the real-time network public opinion data and the user IP address as network public opinion data features; S1-5. Use the page views and platform IP address as auxiliary features of online public opinion data.

[0020] S2 specifically includes: S2-1, obtaining text data features of real-time network public opinion data according to the text data corresponding to the network public opinion data features; S2-2, using the real-time network public opinion data and text data features to establish a basic network public opinion time series data chain according to the analysis reference time t of the network public opinion data features; S2-3. Utilize the basic network public opinion time series data chain to perform verification processing based on time series analysis according to the user IP address of the network public opinion data characteristics to obtain the network public opinion time series data chain.

[0021] S2-1 specifically includes: S2-1-1, using the network public opinion data features to perform text cleaning on the corresponding text data to obtain source text data; S2-1-2, using the source text data to perform word segmentation processing to obtain source text word segmentation data; S2-1-3, using the source text word segmentation data to obtain the word vector of the source text word segmentation data; S2-1-4, establishing a word vector weight for the source text word segmentation data according to the word vector of the source text word segmentation data; S2-1-5, using the word vector and word vector weight of the source text word segmentation data as text data features of real-time network public opinion data; The text cleaning process is to remove noise data in the text.

[0022] In this embodiment, a method for analyzing changes in network public opinion data based on time series analysis is provided. The word vector is established through ont-hot, and the word vector of the text is obtained through basic NLP means. Then, the weight is divided according to the overall text and the noise-reduced data to establish the corresponding word vector weight.

[0023] S2-2 specifically includes: S2-2-1, using the real-time network public opinion data to establish a real-time network public opinion data chain at the analysis reference time t, time t+1 and time t+2 according to the analysis reference time t of the network public opinion data characteristics; S2-2-2, using the text data features to establish a text data feature data chain at the analysis reference time t, time t+1 and time t+2 according to the analysis reference time t of the network public opinion data features; S2-2-3, using the real-time network public opinion data link and the text data feature data link to establish a data-text feature time mapping according to the corresponding time; S2-2-4, obtaining the similarity of the word vectors of the source text segmentation data as the first data link auxiliary feature according to the data-text feature time mapping between the analysis reference time t and the time t+1; S2-2-5, obtaining the similarity of the word vectors of the source text word segmentation data as the second data link auxiliary feature according to the data-text feature time mapping at time t+1 and time t+2; S2-2-6. Use the real-time network public opinion data chain, the text data feature data chain, the first data chain auxiliary feature and the second data chain auxiliary feature as the basic network public opinion time series data chain.

[0024] S2-3 specifically includes: S2-3-1. Based on the user IP address of the network public opinion data characteristics, the user hardware information, the current communication protocol information and the server interface address of the network public opinion data characteristics are obtained based on the time series analysis; S2-3-2, determine whether the user hardware information corresponds to the user IP address, if so, obtain the communication line between the user hardware information and the user IP address, and execute S2-3-3, otherwise, return to S2-3-1; S2-3-3, determine whether the current communication protocol information corresponds to the communication line, if so, obtain the process data corresponding to the communication line and execute S2-3-4, otherwise, return to S2-3-1; S2-3-4, determine whether the process data corresponds to the server interface address, if so, execute S2-3-5, otherwise, return to S2-3-1; S2-3-5, determine whether the number of executions of S2-3-1 at the current moment is greater than 4, if so, abandon the processing, otherwise, retain the current basic network public opinion time series data chain as the network public opinion time series data chain; The communication protocol information at that time is the communication protocol between the user IP address and the server interface address.

[0025] In this embodiment, a method for analyzing changes in network public opinion data based on time series analysis is provided, and the screening basis for the number of executions involved in S2-3-5 is that when S2-3-1 to S2-3-3 are iterated one by one, the normal implementation of the solution can still be guaranteed.

[0026] S3 specifically includes: S3-1, using the network public opinion time series data chain to establish the characteristics of network public opinion data change trends; S3-2. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain and the network public opinion data change trend characteristics.

[0027] S3-1 specifically includes: S3-1-1, using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at the analysis reference time t and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1 to establish a first data change direction; S3-1-2, using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1 and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+2 to establish a second data change direction; S3-1-3, using the network public opinion time series data chain corresponding to the text data feature data chain at the analysis reference time t and the network public opinion time series data chain corresponding to the text data feature data chain at time t+1 to establish a first text feature change direction; S3-1-4, using the text data feature data chain corresponding to the network public opinion time series data chain at time t+1 and the text data feature data chain corresponding to the network public opinion time series data chain at time t+2 to establish a second text feature change direction; S3-1-5, using the network public opinion time series data link to correspond to the first data link auxiliary feature and the second data link auxiliary feature to establish the data link auxiliary feature spacing change direction; S3-1-6. Utilize the first data change direction, the second data change direction, the first text feature change direction, the second text feature change direction and the data link auxiliary feature spacing change direction as the network public opinion data change trend characteristics.

[0028] S3-2 specifically includes: S3-2-1. Use the word vector weight corresponding to the network public opinion time series data chain to establish a data label value; S3-2-2, determine whether the pointing directions of the first data change point and the second data change point are consistent, if so, execute S3-2-3, otherwise, respectively obtain the pointing directions of the data tag value corresponding to the first data change point and the pointing directions of the data tag value corresponding to the second data change point, and execute S3-2-4; S3-2-3, determine whether the average value of the data tag value corresponding to the first data change is consistent with the average value of the data tag value corresponding to the second data change. If so, output the network public opinion time series data chain as the network public opinion data change analysis result. Otherwise, execute S3-2-4; S3-2-4, determine whether the direction of the data tag value corresponding to the first data change is consistent with the direction of the first data change. If so, execute S3-2-5; otherwise, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result, and execute S3-2-5; S3-2-5. Determine whether the direction of the data tag value corresponding to the second data change is consistent with the direction of the second data change. If so, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result. Otherwise, output the real-time network public opinion data corresponding to the analysis base time t, time t+1 and time t+2 as the network public opinion data change analysis result. Among them, the pointing direction includes rising, falling and maintaining, the rising means that the data at the adjacent next moment is greater than the data at the adjacent previous moment, the falling means that the data at the adjacent next moment is less than the data at the adjacent previous moment, and the maintaining means that the data at the adjacent next moment is equal to the data at the adjacent previous moment.

[0029] 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.

[0030] 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.

[0031] 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 including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0032] 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.

[0033] 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 analyzing changes in network public opinion data based on time series analysis, characterized in that: include: S1. Use real-time online public opinion data to obtain online public opinion data characteristics; S2. Establishing a network public opinion time series data chain based on time series analysis according to the real-time network public opinion data and network public opinion data characteristics; S3. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain.

2. The method for analyzing changes in network public opinion data based on time series analysis according to claim 1, characterized in that: The method of using real-time online public opinion data to obtain online public opinion data features includes: Obtain the corresponding time of real-time network public opinion data as the analysis benchmark time t; Obtain text data and user IP address of the real-time network public opinion data; Obtain corresponding page views and platform IP addresses according to the real-time network public opinion data; The analysis reference time t, text data of real-time network public opinion data and user IP address are used as network public opinion data features.

3. The method for analyzing changes in network public opinion data based on time series analysis according to claim 2, characterized in that: Establishing a network public opinion time series data chain based on time series analysis according to the real-time network public opinion data and network public opinion data features includes: S2-1, obtaining text data features of real-time network public opinion data according to the text data corresponding to the network public opinion data features; S2-2, using the real-time network public opinion data and text data features to establish a basic network public opinion time series data chain according to the analysis reference time t of the network public opinion data features; S2-3. Utilize the basic network public opinion time series data chain to perform verification processing based on time series analysis according to the user IP address of the network public opinion data characteristics to obtain the network public opinion time series data chain.

4. The method for analyzing changes in network public opinion data based on time series analysis according to claim 3, characterized in that: Acquiring text data features of real-time network public opinion data according to the network public opinion data features corresponding to the text data includes: Using the network public opinion data features to perform text cleaning on the corresponding text data to obtain source text data; Using the source text data to perform word segmentation processing to obtain source text word segmentation data; Using the source text word segmentation data to obtain the word vector of the source text word segmentation data; Establishing a word vector weight of the source text word segmentation data according to the word vector of the source text word segmentation data; Using the word vectors and word vector weights of the source text word segmentation data as text data features of real-time network public opinion data; The text cleaning process is to remove noise data in the text.

5. The method for analyzing changes in network public opinion data based on time series analysis according to claim 4, characterized in that: Using the real-time network public opinion data and text data features to establish a basic network public opinion time series data chain according to the analysis reference time t of the network public opinion data features includes: Using the real-time network public opinion data, according to the analysis reference time t of the network public opinion data characteristics, a real-time network public opinion data chain at the analysis reference time t, time t+1 and time t+2 is established; Using the text data features, a text data feature data chain at the analysis reference time t, time t+1 and time t+2 is established according to the analysis reference time t of the network public opinion data features; Using the real-time network public opinion data link and the text data feature data link to establish a data-text feature time mapping according to the corresponding time; According to the data-text feature time mapping between the analysis reference time t and the time t+1, the similarity of the word vectors of the source text word segmentation data is obtained as the first data link auxiliary feature; According to the data-text feature time mapping at time t+1 and time t+2, the similarity of the word vectors of the source text word segmentation data is obtained as the second data link auxiliary feature; The real-time network public opinion data chain, the text data feature data chain, the first data chain auxiliary feature and the second data chain auxiliary feature are used as the basic network public opinion time series data chain.

6. The method for analyzing changes in network public opinion data based on time series analysis according to claim 5, characterized in that: The basic network public opinion time series data chain is used to perform verification processing based on the user IP address of the network public opinion data characteristics based on time series analysis to obtain the network public opinion time series data chain, including: S2-3-1. Obtain the user hardware information, current communication protocol information and server interface address of the network public opinion data characteristics based on the time series analysis according to the user IP address of the network public opinion data characteristics; S2-3-2, determine whether the user hardware information corresponds to the user IP address, if so, obtain the communication line between the user hardware information and the user IP address, and execute S2-3-3, otherwise, return to S2-3-1; S2-3-3, determine whether the current communication protocol information corresponds to the communication line, if so, obtain the process data corresponding to the communication line and execute S2-3-4, otherwise, return to S2-3-1; S2-3-4, determine whether the process data corresponds to the server interface address, if so, execute S2-3-5, otherwise, return to S2-3-1; S2-3-5, determine whether the number of executions of S2-3-1 at the current moment is greater than 4, if so, abandon the processing, otherwise, retain the current basic network public opinion time series data chain as the network public opinion time series data chain; The communication protocol information at that time is the communication protocol between the user IP address and the server interface address.

7. The method for analyzing changes in network public opinion data based on time series analysis according to claim 3, characterized in that: The analysis results of network public opinion data changes obtained according to the network public opinion time series data chain include: S3-1, using the network public opinion time series data chain to establish the characteristics of network public opinion data change trends; S3-2. Obtain the results of the network public opinion data change analysis based on the network public opinion time series data chain and the network public opinion data change trend characteristics.

8. The method for analyzing changes in network public opinion data based on time series analysis according to claim 7, characterized in that: Using the network public opinion time series data chain to establish the network public opinion data change trend characteristics includes: Establish a first data change direction by using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at the analysis reference time t and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1; Establish a second data change direction by using the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+1 and the network public opinion time series data chain corresponding to the real-time network public opinion data chain at time t+2; Establish a first text feature change direction by using the text data feature data chain corresponding to the network public opinion time series data chain at the analysis reference time t and the text data feature data chain corresponding to the network public opinion time series data chain at time t+1; Establish a second text feature change direction by using the text data feature data chain corresponding to the network public opinion time series data chain at time t+1 and the text data feature data chain corresponding to the network public opinion time series data chain at time t+2; Using the network public opinion time series data link to correspond to the first data link auxiliary feature and the second data link auxiliary feature, establish a data link auxiliary feature spacing change direction; The first data change direction, the second data change direction, the first text feature change direction, the second text feature change direction and the data link auxiliary feature spacing change direction are used as network public opinion data change trend features.

9. The method for analyzing changes in network public opinion data based on time series analysis according to claim 8, characterized in that: The analysis results of network public opinion data changes obtained based on the network public opinion time series data chain and the network public opinion data change trend characteristics include: S3-2-1. Use the word vector weight corresponding to the network public opinion time series data chain to establish a data label value; S3-2-2, determine whether the pointing directions of the first data change point and the second data change point are consistent, if so, execute S3-2-3, otherwise, respectively obtain the pointing directions of the data tag value corresponding to the first data change point and the pointing directions of the data tag value corresponding to the second data change point, and execute S3-2-4; S3-2-3, determine whether the average value of the data tag value corresponding to the first data change is consistent with the average value of the data tag value corresponding to the second data change. If so, output the network public opinion time series data chain as the network public opinion data change analysis result. Otherwise, execute S3-2-4; S3-2-4, determine whether the direction of the data tag value corresponding to the first data change is consistent with the direction of the first data change. If so, execute S3-2-5; otherwise, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result, and execute S3-2-5; S3-2-5. Determine whether the direction of the data tag value corresponding to the second data change is consistent with the direction of the second data change. If so, output the real-time network public opinion data corresponding to the analysis base time t and time t+1 as the network public opinion data change analysis result. Otherwise, output the real-time network public opinion data corresponding to the analysis base time t, time t+1 and time t+2 as the network public opinion data change analysis result. Among them, the pointing direction includes rising, falling and maintaining, the rising means that the data at the adjacent next moment is greater than the data at the adjacent previous moment, the falling means that the data at the adjacent next moment is less than the data at the adjacent previous moment, and the maintaining means that the data at the adjacent next moment is equal to the data at the adjacent previous moment.

Images