An opinion risk identification method and system based on historical event similarity comparison

Abstract

The application provides a public opinion risk identification method and system based on historical event similarity comparison, which comprises the following steps: determining the field where a target public opinion event is located, and constructing a historical public opinion event knowledge base in the field; using a multi-dimensional public opinion analysis technology to analyze each historical public opinion event in the historical public opinion event knowledge base, and obtaining multi-dimensional analysis results; using a feature similarity algorithm to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and matching the associated events of the target public opinion event; and analyzing the public opinion risk of the target public opinion event based on the multi-dimensional analysis results of the associated events. The application can comprehensively and timely identify the public opinion risk by referring to historical public opinion events, and has high identification efficiency.

Description

Technical Field

[0001] This invention relates to the field of public opinion risk identification technology, and in particular to a method and system for public opinion risk identification based on historical event similarity comparison. Background Technology

[0002] With the rapid development of internet technology and social media applications, various types of online information are experiencing exponential growth. The more information and the greater the traffic, the higher the probability of public opinion risks. In recent years, public opinion crises related to food safety and healthcare have occurred frequently, amplifying irrational and distrustful sentiments in the public sphere and posing significant challenges to government governance and social stability. Against this backdrop, it is crucial to identify and address public opinion risks promptly and accurately, using appropriate technological means, before they escalate (the nascent stage) through specific technical means.

[0003] Referring to current academic and industry methods for identifying public opinion risks, existing technical solutions mainly involve building a public opinion monitoring system, constructing public opinion risk assessment indicators, and supplementing them with human experience-based judgment. This is primarily achieved through the following steps:

[0004] 1) Based on areas of public opinion focus, the public opinion monitoring system can conduct full-network monitoring by setting keywords, or targeted monitoring by locking onto major websites. Currently, there are relatively mature public opinion monitoring technologies and systems on the market, such as Sina Public Opinion Monitoring and Qingbo Public Opinion Monitoring System, which can achieve rapid and convenient early warning of sensitive information, such as real-time push notifications to mobile phones;

[0005] 2) Classify the risk level of discovered public opinion information into high, medium, and low risk using a public opinion risk assessment index system. Existing assessment index systems mainly include the nature of the event involved in the public opinion information, the role of opinion leaders, factors influencing the initial information release, and the public opinion evolution cycle. Specific technical methods mainly include the analytic hierarchy process (AHP), backpropagation (BP) neural network method, accelerated genetic algorithm, UML method, and factor clustering method.

[0006] 3) Based on the information content of monitoring and early warning, the level of public opinion risk, and other information, supplemented by human experience, make public opinion judgments and handling decisions.

[0007] Based on the above steps, it can be seen that existing public opinion risk identification technologies all analyze the actual risk situation of current public opinion events, which is a single perspective. When identifying and judging public opinion risk points, it is necessary to rely on human experience, which is inefficient. Moreover, the public opinion environment is complex and ever-changing. While personal experience can identify some explicit risks, it is not easy to discover potential or hidden public opinion risks in the first instance. Furthermore, the lack of key classification elements and analysis standards makes them of little reference value for public opinion decision-makers in handling public opinion. Summary of the Invention

[0008] In view of this, embodiments of the present invention provide a method and system for identifying public opinion risks based on historical event similarity comparison, so as to eliminate or improve one or more defects existing in the prior art.

[0009] On the one hand, the present invention provides a method for identifying public opinion risks based on historical event similarity comparison, the method comprising the following steps:

[0010] Identify the domain of the target public opinion event and construct a knowledge base of historical public opinion events in that domain;

[0011] Multi-dimensional public opinion analysis technology is used to analyze each historical public opinion event in the historical public opinion event knowledge base to obtain multi-dimensional analysis results;

[0012] A feature similarity algorithm is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and to match the related events of the target public opinion event.

[0013] Based on the multi-dimensional analysis results of the related events, the public opinion risk of the target public opinion event is analyzed.

[0014] In some embodiments of the present invention, the domain of the target public opinion event is determined, and a knowledge base of historical public opinion events in that domain is constructed, including:

[0015] After identifying the area where the target public opinion event is located, collect public opinion data from various sources and in various formats within that area;

[0016] The public opinion data is aggregated into a unified knowledge base and cleaned to construct the historical public opinion event knowledge base.

[0017] In some embodiments of the present invention, multi-dimensional public opinion analysis technology is used to analyze each historical public opinion event in the historical public opinion event knowledge base, including:

[0018] The movement of the parties involved, the movement of key dissemination nodes, the focus of public opinion, the trend of positive and negative public opinion, and the sensitive word cloud are introduced into the public opinion analysis technology. The analysis is presented for each historical public opinion event in the historical public opinion event knowledge base. The public opinion analysis technology includes at least natural language processing, sentiment analysis, and topic modeling.

[0019] In some embodiments of the present invention, the method further includes, regarding the dimension of the movement of the involved parties:

[0020] LDA topic modeling technology is used to extract information about the parties involved from the public opinion data of each historical public opinion event;

[0021] Sentiment analysis techniques are used to measure the tendency of the statements made by the subjects involved, including positive and negative tendencies.

[0022] In some embodiments of the present invention, the method further includes, regarding the dimension of positive and negative public opinion trend:

[0023] The pre-trained BERT model is used to classify the sentiment data of each historical public opinion event and quantify the positive and negative sentiment indices at each moment.

[0024] A public opinion heat map is drawn based on the sentiment index to show the development process of the historical public opinion events and to obtain the trend of public opinion.

[0025] In some embodiments of the present invention, after obtaining the multi-dimensional analysis results, the method further includes:

[0026] The results of the multidimensional analysis are presented in a visual manner, which includes at least time series graphs, network graphs, and word clouds.

[0027] In some embodiments of the present invention, a feature similarity algorithm is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and to match related events of the target public opinion event, including:

[0028] The K-nearest neighbor algorithm is used to calculate the Euclidean distance between the feature vector of the target public opinion event and the feature vectors of each historical public opinion event to obtain the similarity results;

[0029] The historical public opinion events with the highest similarity scores are selected as the related events, sorted from highest to lowest.

[0030] On the other hand, the present invention also provides a public opinion risk identification system based on historical event similarity comparison, the system comprising:

[0031] The data processing module is used to determine the domain of the target public opinion event, construct a knowledge base of historical public opinion events in that domain, and use multi-dimensional public opinion analysis technology to analyze each historical public opinion event in the knowledge base to obtain multi-dimensional analysis results.

[0032] The feature comparison module is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base using a feature similarity algorithm, and to match the related events of the target public opinion event.

[0033] The risk analysis module is used to analyze the public opinion risk of the target public opinion event based on the multi-dimensional analysis results of the related events.

[0034] On the other hand, the present invention provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of any of the methods mentioned above.

[0035] On the other hand, a computer program product includes a computer program / instructions that, when executed by a processor, implement the steps of any of the methods mentioned above.

[0036] This invention provides a method and system for identifying public opinion risks based on historical event similarity comparison, comprising: determining the domain of the target public opinion event and constructing a knowledge base of historical public opinion events in that domain; analyzing each historical public opinion event in the knowledge base using multi-dimensional public opinion analysis technology to obtain multi-dimensional analysis results; comparing the similarity between the target public opinion event and each historical public opinion event in the knowledge base using a feature similarity algorithm to match related events of the target public opinion event; and analyzing the public opinion risk of the target public opinion event based on the multi-dimensional analysis results of related events. The identification method provided by this invention constructs a knowledge base of historical public opinion events in the same field and introduces multi-dimensional public opinion analysis technology to analyze historical public opinion events. It can comprehensively and timely identify public opinion risks by referring to historical public opinion events, assist manual personnel in quickly locating key information points, and identify the explicit or even implicit public opinion risks that the target public opinion event may generate, thereby providing more targeted and comprehensive reference suggestions for public opinion decision-makers. Furthermore, by using a feature similarity algorithm, it can quickly match related events of the target public opinion event without the need for cumbersome search operations on platforms such as Baidu, which greatly improves the identification efficiency.

[0037] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.

[0038] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0039] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, are not intended to limit the scope of the invention. In the drawings:

[0040] Figure 1 This is a schematic diagram illustrating the steps of a public opinion risk identification method based on historical event similarity comparison in one embodiment of the present invention.

[0041] Figure 2 This is a flowchart illustrating a public opinion risk identification method based on historical event similarity comparison in one embodiment of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0043] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0044] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0045] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0046] In the following description, embodiments of the invention will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0047] To address the shortcomings of existing public opinion risk identification technologies, such as a single analytical perspective, reliance on human experience, low identification efficiency, difficulty in timely detection of potential and hidden public opinion risks, and lack of key classification elements and analytical standards, resulting in low reference value, this invention provides a public opinion risk identification method based on historical event similarity comparison. Figure 1 As shown, the method includes the following steps S101 to S104:

[0048] Step S101: Determine the domain of the target public opinion event and construct a knowledge base of historical public opinion events in that domain.

[0049] Step S102: Use multi-dimensional public opinion analysis technology to analyze each historical public opinion event in the historical public opinion event knowledge base to obtain multi-dimensional analysis results.

[0050] Step S103: Use a feature similarity algorithm to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and match the related events of the target public opinion event.

[0051] Step S104: Based on the multi-dimensional analysis results of related events, analyze the public opinion risk of the target public opinion event.

[0052] like Figure 2 The diagram shown is a flowchart of a public opinion risk identification method based on historical event similarity comparison.

[0053] In step S101, the domain of the target public opinion event (the public opinion event to be identified for risk) is determined, and a knowledge base of historical public opinion events in that domain is constructed.

[0054] In some embodiments, public opinion data in the domain of the target public opinion event is collected extensively. This data may originate from various channels such as social media, news websites, blogs, and forums, and may be in diverse formats, including text, images, and audio. The collected public opinion data is aggregated into a unified knowledge base and then cleaned and optimized to construct a historical public opinion event knowledge base for that domain. The preprocessing, including cleaning, of the public opinion data aims to remove noisy and duplicate data, standardize the content from different data sources, and prepare for subsequent analysis.

[0055] In some embodiments, the above method can be used to build a knowledge base of historical public opinion events in multiple fields in advance. After determining the field in which the target public opinion event is located, the knowledge base of historical public opinion events in the corresponding field can be directly called.

[0056] In step S102, multi-dimensional public opinion analysis technology is used to analyze each historical public opinion event in the historical public opinion event knowledge base to obtain multi-dimensional analysis results.

[0057] In some embodiments, five dimensions—the movements of the parties involved, the movements of key dissemination nodes, the focus of public opinion, the trends of positive and negative public opinion, and a sensitive word cloud—are incorporated into public opinion analysis technology. This data is then analyzed and presented for each historical public opinion event within a historical public opinion event knowledge base. Commonly used public opinion analysis techniques include natural language processing, sentiment analysis, and topic modeling.

[0058] Taking the movements of the parties involved as an example: The parties involved refer to key individuals, organizations, or institutions that are directly or indirectly involved in a public opinion event. These parties may include enterprises, government departments, celebrities, social media influencers, public figures, etc.

[0059] Analyzing the behavior of the parties involved helps us understand how they change during an event and how their statements or actions influence public sentiment and opinion. For example, a company's statement on an event might be negative or ambiguous, which could exacerbate public opinion; conversely, an organization's positive response could alleviate public pressure.

[0060] LDA topic modeling technology can be used to extract information about the subjects involved from the public opinion data of each historical public opinion event, and then sentiment analysis technology can be used to measure whether the subjects' opinions are positive or negative.

[0061] Taking the trend of positive and negative public opinion heat as an example: The trend of positive and negative public opinion heat refers to the fluctuation of public sentiment in a public opinion event, including the changing trend of positive emotions (such as support and praise) and negative emotions (such as anger and criticism).

[0062] Pre-trained deep learning models such as BERT and LSTM can be used to classify the sentiment data of each historical public opinion event into positive, negative, and neutral sentiments, and quantify the positive and negative sentiment indices at each moment. Based on the sentiment indices, a public opinion heat map can be drawn to track the development process of historical public opinion events, analyze the accumulation and spread of positive or negative sentiments, and thus obtain the trend of public opinion.

[0063] In some embodiments, after obtaining the multi-dimensional analysis results, the results are presented in a visual manner, such as time series graphs, network graphs, and word clouds. Time series graphs can show the changing trends of sentiment and public opinion heat, network graphs can show key dissemination nodes and dissemination paths, and word clouds can show sensitive words and public attention.

[0064] In step S103, a feature similarity algorithm is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and to match the related events of the target public opinion event.

[0065] In some embodiments, the target public opinion event's attribute features are decomposed to construct a feature vector, such as text features (e.g., keywords, topic words, sentiment polarity, etc.), social dissemination features (e.g., number of comments, number of reposts, dissemination speed, etc.), and temporal features (e.g., occurrence time, duration, etc.). The K-nearest neighbor algorithm is used to compare the feature vector of the target public opinion event with the feature vectors of each historical public opinion event in the historical public opinion event knowledge base, calculating the Euclidean distance between them to obtain the similarity result. The similarity calculation method can not only calculate Euclidean distance but also cosine similarity, Manhattan distance, etc. The similarity results are sorted from high to low, and the historical public opinion event with the highest similarity result is selected as the associated event of the target public opinion event.

[0066] In step S104, based on the multi-dimensional analysis results of step S102, key information of related events (historical public opinion events with the highest similarity to the target public opinion event) is obtained. By analyzing the key information of related events, the public opinion risks that the target public opinion event may face are predicted, such as the spread trend of the target public opinion event, public reaction, and possible positive and negative public opinion risks, so as to discover possible risk points in advance and take intervention measures or formulate response strategies in advance.

[0067] Corresponding to the public opinion risk identification method based on historical event similarity comparison, the present invention also provides a public opinion risk identification system based on historical event similarity comparison, the system comprising:

[0068] The data processing module is used to determine the domain of the target public opinion event, construct a knowledge base of historical public opinion events in that domain, and use multi-dimensional public opinion analysis technology to analyze each historical public opinion event in the knowledge base to obtain multi-dimensional analysis results.

[0069] The feature comparison module is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base using a feature similarity algorithm, and to match the related events of the target public opinion event.

[0070] The risk analysis module is used to analyze the public opinion risks of a target public opinion event based on the multi-dimensional analysis results of related events.

[0071] Corresponding to the above method, the present invention also provides an electronic device, which includes a computer device, the computer device including a processor and a memory, the memory storing computer instructions, the processor executing the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the electronic device performs the steps of the method as described above.

[0072] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the aforementioned edge computing server deployment method. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.

[0073] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0074] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0075] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0076] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying public opinion risks based on historical event similarity comparison, characterized in that, The method includes the following steps: Identify the domain of the target public opinion event and construct a knowledge base of historical public opinion events in that domain; Multi-dimensional public opinion analysis technology is used to analyze each historical public opinion event in the historical public opinion event knowledge base to obtain multi-dimensional analysis results; A feature similarity algorithm is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and to match the related events of the target public opinion event. Based on the multi-dimensional analysis results of the related events, the public opinion risk of the target public opinion event is analyzed.

2. The public opinion risk identification method based on historical event similarity comparison according to claim 1, characterized in that, Identify the domain of the target public opinion event and construct a knowledge base of historical public opinion events in that domain, including: After identifying the area where the target public opinion event is located, collect public opinion data from various sources and in various formats within that area; The public opinion data is aggregated into a unified knowledge base and cleaned to construct the historical public opinion event knowledge base.

3. The public opinion risk identification method based on historical event similarity comparison according to claim 1, characterized in that, Multi-dimensional public opinion analysis techniques are used to analyze each historical public opinion event in the historical public opinion event knowledge base, including: The movement of the parties involved, the movement of key dissemination nodes, the focus of public opinion, the trend of positive and negative public opinion, and the sensitive word cloud are introduced into the public opinion analysis technology. The analysis is presented for each historical public opinion event in the historical public opinion event knowledge base. The public opinion analysis technology includes at least natural language processing, sentiment analysis, and topic modeling.

4. The public opinion risk identification method based on historical event similarity comparison according to claim 3, characterized in that, Regarding the dimension of the movements of the parties involved, the method further includes: LDA topic modeling technology is used to extract information about the parties involved from the public opinion data of each historical public opinion event; Sentiment analysis techniques are used to measure the tendency of the statements made by the subjects involved, including positive and negative tendencies.

5. The public opinion risk identification method based on historical event similarity comparison according to claim 3, characterized in that, Regarding the aforementioned positive and negative public opinion trend dimensions, the method further includes: The pre-trained BERT model is used to classify the sentiment data of each historical public opinion event and quantify the positive and negative sentiment indices at each moment. A public opinion heat map is drawn based on the sentiment index to show the development process of the historical public opinion events and to obtain the trend of public opinion.

6. The method for identifying public opinion risks based on historical event similarity comparison according to claim 1, characterized in that, After obtaining the multidimensional analysis results, it also includes: The results of the multidimensional analysis are presented in a visual manner, which includes at least time series graphs, network graphs, and word clouds.

7. The method for identifying public opinion risks based on historical event similarity comparison according to claim 1, characterized in that, A feature similarity algorithm is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base, and to match the related events of the target public opinion event, including: The K-nearest neighbor algorithm is used to calculate the Euclidean distance between the feature vector of the target public opinion event and the feature vectors of each historical public opinion event to obtain the similarity results; The historical public opinion events with the highest similarity scores are selected as the related events, sorted from highest to lowest.

8. A public opinion risk identification system based on historical event similarity comparison, characterized in that, The system includes: The data processing module is used to determine the domain of the target public opinion event, construct a knowledge base of historical public opinion events in that domain, and use multi-dimensional public opinion analysis technology to analyze each historical public opinion event in the knowledge base to obtain multi-dimensional analysis results. The feature comparison module is used to compare the similarity between the target public opinion event and each historical public opinion event in the historical public opinion event knowledge base using a feature similarity algorithm, and to match the related events of the target public opinion event. The risk analysis module is used to analyze the public opinion risk of the target public opinion event based on the multi-dimensional analysis results of the related events.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method as described in any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method as described in any one of claims 1 to 7.

Images