Viewpoint evolution analysis method and system for social platform hot events, electronic equipment and storage medium

Abstract

The invention discloses a viewpoint evolution analysis method and system for a social platform hotspot event, electronic equipment and a storage medium. The method comprises the following steps: collecting speech data about a specific target on a social platform; performing time sequence analysis on the speech data, dividing the evolution process of the viewpoint into a plurality of time periods, and identifying a plurality of effective wave crests discussed by a specific target; and based on the speech post in each effective wave crest time period, extracting a keyword of the time period, and generating a representative viewpoint of the time period. Through time sequence analysis and wave crest detection, viewpoint evolution processes of events or topics at different time stages can be displayed in a fine-grained manner, so that a user can clearly know how viewpoints evolve along with time change, the peak period discussed by the user can be accurately positioned, representative viewpoints are extracted from the peak period, and the user experience is improved. And the user is helped to grasp the key evolution moment, so that the dynamic evolution of a specific target or topic viewpoint is more efficiently understood.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method and system for analyzing the evolution of opinions on hot events on a social platform, an electronic device, and a computer-readable storage medium. Background Art

[0002] With the rapid development of the Internet and social media, social platforms have become one of the important channels for information dissemination. Users express their views, exchange opinions, and spread information through social platforms. However, the rapid changes in information flow and the explosive growth of information volume on social platforms have brought great challenges to users in obtaining key information and understanding the full picture of events. In hot events, tens of thousands of users comment on the same event, and these opinions often undergo multiple changes and evolutions during the development of the event, especially in different time periods. Form representative high-frequency discussion points or opinion "peaks". Therefore, how to effectively extract these opinions and analyze their evolution over time has become a technical problem.

[0003] At present, existing social data analysis methods mainly focus on the following aspects:

[0004] 1. Keyword extraction and sentiment analysis: Existing technologies usually rely on keyword extraction and sentiment analysis to process social media data. Such methods are based on text mining and extract event-related keywords through word frequency analysis, TF-IDF (TermFrequency-Inverse Document Frequency), latent semantic analysis (LSA) and other methods of social media texts, and identify the emotional tendencies of user speeches through sentiment classification algorithms. Such methods can provide basic emotional information and hot keywords for some events, but they are difficult to fully display the evolution of opinions at different stages.

[0005] 2. Information clustering and topic detection: Another type of technology mainly uses clustering and topic models to aggregate massive social data in order to identify some major topics. Common clustering techniques include K-means clustering, hierarchical clustering, LDA (latent Dirichlet allocation) topic model, etc. These methods can aggregate comments with similar content to obtain a certain number of major discussion topics. However, this type of method mainly focuses on static topic extraction and cannot well capture and present the changes and dynamic evolution of opinions in different time periods during the development of events.

[0006] 3. Time series analysis: In view of the dynamic characteristics of information on social media, existing technologies have also tried to introduce time series analysis to describe the changing trends of social media data in different time periods. For example, data sampling is performed using a time window method, and the content within each time window is analyzed to observe the changes in user opinions. However, such methods often have difficulty in accurately identifying important time nodes, that is, the time period when user discussion heat increases significantly during the development of an event.

[0007] In summary, existing social data analysis methods have the following deficiencies in analyzing opinions and comments on social platforms:

[0008] 1. Lack of fine-grained presentation of the evolution of opinions: Although existing keyword extraction, topic clustering and sentiment analysis methods can identify the main content and emotional changes of events to a certain extent, most of these methods remain at the overall level and cannot clearly show the formation, development and evolution of event-related opinions in different time periods. This makes it difficult for users to fully understand the development context of the event, especially when the event is complex and the opinions are diverse. It is impossible to accurately show the correlation and evolution trend between different opinions.

[0009] 2. Difficulty in accurately identifying key nodes: Existing time series analysis methods often rely on fixed time windows to segment data, making it difficult to automatically identify key nodes in an event (i.e., the "peaks" of discussion heat), and thus unable to locate the moments when opinions change significantly during the development of an event.

[0010] 3. Information redundancy and repetition: Social media is filled with a large amount of similar or repetitive content. Existing analysis methods are insufficient in removing information redundancy, resulting in the analysis results often being unable to refine the most representative and innovative viewpoints, affecting users' effective understanding of events and grasp of core information. Summary of the invention

[0011] The present invention provides a method and system for analyzing the evolution of opinions on hot events on a social platform, an electronic device, and a computer-readable storage medium, which can display the evolution process of opinions on events or topics at different time stages in a fine-grained manner, allowing users to clearly understand how opinions evolve over time, and can accurately locate the peak periods of user discussions and extract representative opinions therefrom, helping users grasp the key evolution moments of events, thereby more efficiently understanding the dynamic evolution of events.

[0012] According to one aspect of the present invention, a method for analyzing the evolution of opinions on hot events on a social platform is provided, comprising the following contents:

[0013] Collect speech data about specific targets on social platforms;

[0014] Conduct time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions;

[0015] Based on the speech posts in each effective peak time period, the keywords of the time period are extracted, and the representative opinions of the time period are generated.

[0016] Furthermore, the process of performing time series analysis on speech data, dividing the evolution of opinions into multiple time periods, and identifying multiple effective peak periods of specific target discussions includes the following:

[0017] The speech data is grouped by day, and the number of discussions per day is calculated. All discussion peaks are identified by analyzing the rising and falling trends of the discussion numbers. The number of discussions at each discussion peak is compared with the discussion number threshold. If the number of discussions at a discussion peak exceeds the discussion number threshold, the discussion peak is regarded as a valid peak.

[0018] Furthermore, after all discussion peaks are identified, the highest peak and the lowest peak are removed, and the average value of the discussion numbers of the remaining discussion peaks is calculated, and the average value is used as the discussion number threshold.

[0019] Furthermore, after all valid peaks are identified, a merging operation is performed on continuous parallel peaks with the same peak value and valid peaks with overlapping time.

[0020] Furthermore, after collecting speech data about a specific target on a social platform, the following contents are also included:

[0021] Clean and format speech data.

[0022] Furthermore, after all representative viewpoints are generated, the following contents are also included:

[0023] Perform similarity checks on all representative opinions and filter out duplicate or highly similar opinions.

[0024] Furthermore, after all representative viewpoints are generated, the following contents are also included:

[0025] Visualize the results of opinion evolution analysis.

[0026] In addition, the present invention also provides a system for analyzing the evolution of opinions on hot events on a social platform, including:

[0027] The data collection module is used to collect speech data about specific targets on social platforms;

[0028] The time series analysis module is used to perform time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions;

[0029] The keyword extraction and opinion generation module is used to extract the keywords of each effective peak time period based on the speech posts in the time period and generate representative opinions of the time period.

[0030] In addition, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0031] In addition, the present invention also provides a computer-readable storage medium for storing a computer program for analyzing the evolution of opinions on hot events on social platforms, and the computer program executes the steps of the method described above when running on a computer.

[0032] The present invention has the following beneficial effects:

[0033] The method for analyzing the evolution of opinions on hot events on social platforms of the present invention, after collecting speech data on a designated event or topic on a social platform, divides the target's evolution process into multiple time periods by performing time series analysis on the speech data, and identifies multiple effective peaks of target discussions, and then extracts the keywords of the time period based on the speech posts in each effective peak time period, and generates representative opinions of the time period. Through time series analysis and peak detection, the method can display the opinion evolution process of an event or topic at different time stages in a fine-grained manner, so that users can clearly understand how opinions evolve over time, overcome the deficiency of the prior art that the evolution of opinions cannot be deeply displayed, and can accurately locate the peak period of user discussion, and extract representative opinions from it, helping users grasp the key evolution moments of events, thereby more efficiently understanding the dynamic evolution of events, and significantly improving the accuracy and effectiveness of event analysis on social platforms, helping users better understand the development context of events and the dynamic evolution of their core opinions.

[0034] In addition, the opinion evolution analysis system for hot events on social platforms of the present invention also has the above advantages.

[0035] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0037] Figure 1 It is a flowchart of a method for analyzing the evolution of opinions on hot events on a social platform according to a preferred embodiment of the present application;

[0038] Figure 2 It is another flow chart of the method for analyzing the evolution of opinions on hot events on social platforms according to a preferred embodiment of the present application;

[0039] Figure 3 It is another flow chart of the method for analyzing the evolution of opinions on hot events on social platforms according to a preferred embodiment of the present application;

[0040] Figure 4 It is another flow chart of the method for analyzing the evolution of opinions on hot events on social platforms according to a preferred embodiment of the present application;

[0041] Figure 5 It is a schematic diagram of the module structure of a system for analyzing the evolution of opinions on hot events on a social platform according to another embodiment of the present application. DETAILED DESCRIPTION

[0042] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0043] Reference Figure 1 As shown, the preferred embodiment of the present application provides a method for analyzing the evolution of opinions on hot events on a social platform, including the following contents:

[0044] Step S1: Collect speech data about a specific target on the social platform;

[0045] It can be understood that in step S1, the specific target refers to a specific event or topic. When the present invention detects a new hot event or topic, it will extract the keywords therein, and then use the keywords to collect raw data on the social platform. Specifically, for example, the specific target is event A, and its keywords are a, b, and c. The present invention interacts with the API interface of the social platform and uses the keywords a, b, and c to periodically capture speech data related to the event on the social platform, wherein the speech data includes the document of the speech post, the user's posting, likes, comments, forwarding, citation and other behavioral data. Posting refers to the speech published by the user on his own social homepage, likes refer to the likes of other users' speeches, comments refer to the user's comments on other users' speeches, forwarding refers to the user directly posting other users' speeches on his own social homepage, and quoting refers to the user posting other users' speeches on his own social homepage with his own speech. These speech data provide the original data source for subsequent data processing and analysis, ensuring the comprehensiveness and timeliness of the data, so as to ensure that the data analyzed subsequently can cover all aspects of the event.

[0046] Step S2: Perform time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions;

[0047] It can be understood that in step S2, the process of performing time series analysis on speech data, dividing the target evolution process into multiple time periods, and identifying multiple effective peak periods of target discussion includes the following:

[0048] The speech data is grouped by day, and the number of discussions per day is calculated. All discussion peaks are identified by analyzing the rising and falling trends of the discussion numbers. The number of discussions at each discussion peak is compared with the discussion number threshold. If the number of discussions at a discussion peak exceeds the discussion number threshold, the discussion peak is regarded as a valid peak.

[0049] Specifically, first, the speech data is grouped by day, and the number of discussions per day is calculated. All discussion peaks are identified by analyzing the upward and downward trends of the discussion numbers. Based on all discussion peaks, the discussion process of the entire event can be divided into multiple stages. Secondly, the highest peak and the lowest peak are removed, and the average number of discussions of the remaining discussion peaks is calculated, and the average value is used as the discussion number threshold. Then, the number of discussions of all discussion peaks is compared with the discussion number threshold. If the number of discussions of a discussion peak exceeds the discussion number threshold, the discussion peak is used as a valid peak. Finally, after all valid peaks are identified, continuous parallel peaks with the same peak value and valid peaks with overlapping time are merged to make the detection result of valid peaks more accurate.

[0050] It can be understood that the present invention can display the evolution process of opinions on events or topics at different time stages in a fine-grained manner through time series analysis and peak detection, so that users can clearly understand how opinions evolve over time, and can accurately locate the peak periods of user discussions, helping users grasp the key evolution moments of events, thereby understanding the dynamic evolution of events more efficiently.

[0051] Optionally, a sliding time window can be used instead of a fixed day grouping, so that the window size can be dynamically adjusted according to the amount of data, and more flexibly adapt to the discussion characteristics of different events. In addition, in addition to peak detection based on the amount of discussion, multiple dimensions such as user activity and emotional intensity can be introduced for comprehensive peak determination. In addition, neural networks or other machine learning models can be used to predict possible discussion peaks and identify key time nodes in advance. In addition, time series analysis can be performed on different user groups separately to obtain more detailed characteristics of the evolution of group opinions.

[0052] Step S3: extract the keywords of each effective peak time period based on the speech posts in the time period, and generate representative opinions of the time period.

[0053] It can be understood that in step S3, after determining the effective peak time period, the speech posts in each effective peak time period are deeply analyzed and processed, the keywords of the time period are extracted and the representative views of the time period are generated. Specifically, the speech posts in each effective peak time period are sorted in chronological order, and the keywords are extracted by combining the large language model with traditional keyword extraction. In addition, the historical keywords before the peak can be collected as a benchmark reference, which is conducive to accurately generating new views. Among them, the large language model is pre-fine-tuned and trained with a certain number of high-quality documents, which can ensure the quality of keyword extraction. When the number of documents is large, the traditional keyword extraction method is used for assistance. Through this hybrid strategy, the accuracy and representativeness of keyword extraction can be guaranteed, and the processing efficiency can be ensured. After extracting the keywords, the keywords are combined with other attribute information of the speech posts, such as the theme, time, person entity and institution entity, etc., to construct a complete representative viewpoint, and group and organize them by time range and topic tags.

[0054] It can be understood that the present invention, by combining a large language model and a traditional keyword extraction algorithm, can identify representative discussion focuses from a large amount of data and generate concise and clear descriptions of opinions. At the same time, representative opinions are displayed in a time series manner, which can well reflect the evolution path of opinions.

[0055] Optionally, you can also use pre-trained language models such as BERT (Bidirectional Encoder Representations from Transformers) to replace traditional keyword extraction methods. BERT can more accurately extract the core information of the text through contextual relationships, thereby improving the quality and representativeness of opinion generation.

[0056] It can be understood that the method for analyzing the evolution of opinions on hot events on social platforms of this embodiment, after collecting speech data on a designated event or topic on a social platform, divides the target's evolution process into multiple time periods by performing time series analysis on the speech data, and identifies multiple effective peaks of target discussions, and then extracts the keywords of the time period based on the speech posts in each effective peak time period, and generates representative opinions of the time period. Through time series analysis and peak detection, this method can display the evolution of opinions on events or topics at different time stages in a fine-grained manner, so that users can clearly understand how opinions evolve over time, overcome the deficiency of the prior art that it is not possible to deeply display the evolution of opinions, and can accurately locate the peak period of user discussions, and extract representative opinions from them, helping users grasp the key evolution moments of events, so as to more efficiently understand the dynamic evolution of events, and can significantly improve the accuracy and effectiveness of event analysis on social platforms, and help users better understand the development context of events and the dynamic evolution of their core opinions.

[0057] like Figure 2 As shown, the method for analyzing the evolution of opinions on hot events on social platforms further includes the following contents after collecting speech data:

[0058] Step S1a: Clean and format speech data.

[0059] Specifically, after collecting speech data, the speech data is cleaned and formatted, including removing noise data (such as advertisements and irrelevant content), removing duplicate content, and converting the data into a unified format for subsequent analysis. This not only ensures the quality and consistency of the data, but also effectively filters out irrelevant or low-quality data, reduces redundant information in the data, and makes subsequent analysis more accurate and efficient, thereby improving the accuracy of analysis.

[0060] In addition, after the speech data is cleaned and formatted, it is stored in a distributed database to support efficient query and analysis operations, ensuring that all cleaned and formatted data can be persistently stored. This data storage method not only improves the security of the data, but also provides an efficient access path for subsequent time series analysis and keyword extraction. The use of a distributed database ensures that the present invention can process a large amount of speech data and quickly retrieve data when needed. In addition, in the case of small-scale data processing, a simple data storage method can be used instead of a distributed database, or analysis can be performed directly after collection. However, when processing large-scale speech data, distributed storage can improve the performance of the present invention, enhance scalability and reliability, and support the efficiency and flexibility of subsequent analysis.

[0061] like Figure 3 As shown, the method for analyzing the evolution of opinions on hot events on social platforms, after generating all representative opinions, further includes the following contents:

[0062] Step S4: Perform similarity detection on all representative opinions and filter out duplicate or highly similar opinions.

[0063] Specifically, by comparing the newly generated keywords with the historical keywords, duplicate or highly similar opinions are filtered out. Among them, the specific similarity evaluation method can use the existing cosine similarity algorithm, for example, vector encoding the keywords, and then calculating the cosine value of the angle between the vector corresponding to the newly generated keyword and the vector corresponding to the historical keyword. If the cosine value of the angle is greater than the preset threshold, it is determined that the two are highly similar. If the cosine value of the angle is 1, it is determined that the two are repeated. In addition, necessary attribute information, such as user information, time, etc., can be supplemented for each representative opinion to provide complete contextual support. In addition, for each effective peak time period, the most representative opinion will be selected as the main display content, while other opinions will be retained as supplementary information. In addition, a text similarity algorithm can be used to replace the cosine similarity calculation. For example, a text embedding model based on deep learning (such as Word2Vec or Sentence-BERT) can represent text features in a higher dimension, thereby improving the accuracy of similarity detection and reducing redundant information.

[0064] It can be understood that the present invention adopts a deduplication mechanism of similarity detection to remove duplicate or overly similar viewpoints, thereby ensuring that the ultimately generated viewpoints are as independent and representative as possible in content, which can effectively reduce redundant information, improve the accuracy and representativeness of analysis results, and effectively improve the output quality of viewpoints, thereby avoiding users from encountering a large amount of duplicate information when viewing results, and ensuring that what is ultimately presented to users is a set of viewpoints that is both temporal, phased, and representative.

[0065] In addition, if Figure 4 As shown, the method for analyzing the evolution of opinions on hot events on social platforms, after generating all representative opinions, further includes the following contents:

[0066] Step S5: Visualize and output the results of opinion evolution analysis.

[0067] Specifically, after similarity detection and deduplication, the final results of the opinion evolution analysis are presented to users through the user interface. The user interface displays the evolution of opinions in the form of visual charts, including the main opinions in each time period, the peak period of discussion, and the changing trend of opinions. Users can further view the detailed information of each stage through interactive functions to help users better understand the dynamic changes and key opinions of events, greatly improving the efficiency of users' understanding of complex events.

[0068] It can be understood that the opinion evolution analysis method of hot events on social platforms of the present invention can be applied to a variety of social platform event analysis scenarios, such as analyzing the evolution of opinions on large-scale public events, tracking changes in comments on social platforms, monitoring public evaluations of brands, etc. In these scenarios, the present invention can automatically identify the peak moments of discussion, extract key opinions, and display the evolution path of opinions in a time series manner.

[0069] like Figure 5 As shown, another embodiment of the present invention further provides a system for analyzing the evolution of opinions on hot events on a social platform, preferably using the above-mentioned method for analyzing the evolution of opinions on hot events on a social platform, including:

[0070] The data collection module is used to collect speech data about specific targets on social platforms;

[0071] The time series analysis module is used to perform time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions;

[0072] The keyword extraction and opinion generation module is used to extract the keywords of each effective peak time period based on the speech posts in the time period and generate representative opinions of the time period.

[0073] It can be understood that the opinion evolution analysis system of hot events on social platforms of this embodiment, after collecting speech data about a specified event or topic on a social platform, divides the target's evolution process into multiple time periods by performing time series analysis on the speech data, and identifies multiple effective peaks of target discussions, and then extracts the keywords of the time period based on the speech posts in each effective peak time period, and generates representative opinions of the time period. Through time series analysis and peak detection, the system can display the opinion evolution process of events or topics at different time stages in a fine-grained manner, so that users can clearly understand how opinions evolve over time, overcome the deficiency of the prior art that it is unable to deeply display the evolution of opinions, and can accurately locate the peak period of user discussions, and extract representative opinions from them, helping users grasp the key evolution moments of events, so as to more efficiently understand the dynamic evolution of events, which can significantly improve the accuracy and effectiveness of event analysis on social platforms, and help users better understand the development context of events and the dynamic evolution of their core opinions.

[0074] In addition, the social platform hot event opinion evolution analysis system also includes:

[0075] The data preprocessing module is used to clean and format speech data.

[0076] In addition, the social platform hot event opinion evolution analysis system also includes:

[0077] The similarity detection module is used to perform similarity detection on all representative opinions and filter out repeated or highly similar opinions.

[0078] In addition, the social platform hot event opinion evolution analysis system also includes:

[0079] The visualization display module is used to visualize the results of opinion evolution analysis.

[0080] It can be understood that the various modules of the system embodiment correspond to the various steps of the above method embodiment, so the specific working principles of each module are not repeated here, and the corresponding references can be made to the various steps of the above method embodiment.

[0081] In addition, another embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0082] In addition, another embodiment of the present invention further provides a computer-readable storage medium for storing a computer program for analyzing the evolution of opinions on hot events on social platforms, wherein the computer program executes the steps of the method described above when running on a computer.

[0083] Common computer-readable storage media include: floppy disks, flexible disks, hard disks, magnetic tapes, any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tapes, any other physical media with patterns of holes, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash-erasable programmable read-only memory (FLASH-EPROM), any other memory chip or cartridge, or any other medium that can be read by a computer. Instructions can further be transmitted or received by a transmission medium. The term transmission medium may include any tangible or intangible medium that can be used to store, encode or carry instructions for execution by a machine, and includes digital or analog communication signals or intangible media that facilitate communication of the above instructions. Transmission media include coaxial cables, copper wire, and optical fiber, including the wires of a bus used to transmit a computer data signal.

[0084] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt 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 codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0085] The present application 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 box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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.

[0086] 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.

[0087] 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 in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0088] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0089] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0090] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for analyzing the evolution of opinions on hot events on social platforms, characterized in that: Includes the following: Collect speech data about specific targets on social platforms; Conduct time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions; Based on the speech posts in each effective peak time period, the keywords of the time period are extracted, and the representative opinions of the time period are generated.

2. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 1, characterized in that: The process of performing time series analysis on speech data, dividing the evolution of opinions into multiple time periods, and identifying multiple effective peak periods of specific target discussions includes the following: The speech data is grouped by day, and the number of discussions per day is calculated. All discussion peaks are identified by analyzing the rising and falling trends of the discussion numbers. The number of discussions at each discussion peak is compared with the discussion number threshold. If the number of discussions at a discussion peak exceeds the discussion number threshold, the discussion peak is regarded as a valid peak.

3. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 2, characterized in that: After all discussion peaks are identified, the highest peak and the lowest peak are removed, and the average discussion number of the remaining discussion peaks is calculated, and the average is used as the discussion number threshold.

4. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 2, characterized in that: After all valid peaks are identified, continuous parallel peaks with the same peak value and valid peaks with overlapping time are merged.

5. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 1, characterized in that: After collecting speech data about specific targets on social platforms, the following contents are also included: Clean and format speech data.

6. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 1, characterized in that: After generating all representative views, Includes the following: Perform similarity checks on all representative opinions and filter out duplicate or highly similar opinions.

7. The method for analyzing the evolution of opinions on hot events on social platforms as claimed in claim 1, characterized in that: After generating all representative views, Includes the following: Visualize the results of opinion evolution analysis.

8. A system for analyzing the evolution of opinions on hot events on social platforms, characterized in that: include: The data collection module is used to collect speech data about specific targets on social platforms; The time series analysis module is used to perform time series analysis on speech data, divide the evolution of opinions into multiple time periods, and identify multiple effective peaks of specific target discussions; The keyword extraction and opinion generation module is used to extract the keywords of each effective peak time period based on the speech posts in the time period and generate representative opinions of the time period.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the method according to any one of claims 1 to 7 by calling the computer program stored in the memory.

10. A computer-readable storage medium for storing a computer program for analyzing the evolution of opinions on hot events on a social platform, characterized in that: When the computer program is run on a computer, the steps of the method according to any one of claims 1 to 7 are executed.

Images