Dynamic network internet emotion analysis method and system based on sample expansion

Abstract

The invention discloses a dynamic network Internet emotion analysis method and system based on sample expansion, and relates to the technical field of emotion analysis based on a graph network, the method comprises the following steps: constructing an emotion propagation graph network, firstly collecting emotion data of a user from the Internet, and mapping the collected emotion data to a graph network; in the constructed graph network, further calculating the intensity of emotion propagation through an algorithm; a future emotion change trend of a specific group is predicted by establishing a sequential relationship of emotion propagation, and abnormal emotion fluctuation is monitored in real time; and results of emotion trend prediction and anomaly detection are fed back to the emotion propagation graph network, and a sample expansion strategy is optimized. By combining the emotion trend and the anomaly detection result, the invention provides a comprehensive emotion propagation evaluation mechanism, which is helpful for decision makers to make more accurate prediction and response measures. The method has important application value for advertisement marketing and public opinion management.

Description

Technical Field

[0001] The present invention relates to the technical field of sentiment analysis based on graph networks, and in particular to a dynamic network Internet sentiment analysis method and system based on sample augmentation. Background Art

[0002] In recent years, with the rapid development of the Internet and social media, the emotional expressions of users on network platforms have gradually become important data sources in the fields of social research, marketing analysis, and public safety. Sentiment analysis technology helps all parties understand public sentiment, trend changes, and potential social impacts by mining and analyzing users' online behaviors, social interactions, and emotional expressions. Therefore, how to accurately and efficiently analyze the patterns of emotional propagation and perform dynamic prediction and anomaly detection based on this has become an important issue in the field of sentiment analysis.

[0003] Most current sentiment analysis methods are based on a single sentiment dataset, usually static sentiment analysis, lacking in-depth mining and real-time monitoring of the emotional propagation process. In the prior art, some methods construct graph networks using nodes (such as users) and edges (such as relationships) in social networks to analyze the paths and scopes of emotional propagation. However, these methods often ignore the temporal and dynamic characteristics of emotional propagation, especially the weak early warning and response capabilities for emotional outbreaks (such as cyber violence, public opinion crises, etc.). Although some studies have tried to introduce time factors in emotional propagation analysis, they lack effective temporal prediction and dynamic adjustment mechanisms, resulting in poor prediction accuracy and real-time performance.

[0004] Existing sample augmentation methods are mostly based on static data enhancement methods, lacking detailed modeling of the propagation paths and their intensity changes, and failing to effectively integrate the temporal change characteristics in the graph network structure. In the process of dynamic emotional propagation analysis, how to dynamically adjust the propagation intensity and scope from the time dimension and how to identify and early warn abnormal emotional fluctuations are still a major challenge in sentiment analysis. Summary of the Invention

[0005] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the specification of this application, to avoid obscuring the purpose of this part, the abstract, and the title. However, such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the problems existing in the above-mentioned prior dynamic network Internet sentiment analysis method and system based on sample augmentation, the present invention is proposed.

[0007] To solve the above technical problems, the present invention provides the following technical solutions:

[0008] A dynamic network Internet sentiment analysis method based on sample augmentation, the method comprising the following steps:

[0009] Step 1: Construct a sentiment propagation graph network. First, collect users' sentiment data from the Internet and map the collected sentiment data into a graph network;

[0010] Among them, the nodes i, j of the graph network represent users or events; the edges represent the relationships between users; the initial weight S of the nodes i represents the intensity of the sentiment expressed by the user;

[0011] Step 2: In the constructed graph network, further calculate the intensity of sentiment propagation through an algorithm. Through the analysis of the propagation path and intensity, key influencing nodes can be screened out;

[0012] Step 3: Utilize the propagation path and intensity data obtained in Step 2 to predict the future sentiment change trend of a specific group by establishing the temporal relationship of sentiment propagation. When predicting the sentiment change trend, monitor abnormal sentiment fluctuations in real time;

[0013] Step 4: Feed the results of sentiment trend prediction and anomaly detection back into the sentiment propagation graph network to further optimize the sample augmentation strategy.

[0014] As a preferred solution of the dynamic network Internet sentiment analysis method based on sample augmentation according to the present invention, wherein: in the said Step 2, the steps of calculating the intensity and range of sentiment propagation specifically include:

[0015] S201: According to the sentiment intensity difference ΔS between nodes i and j ij (t) = S i (t) - S j (t), calculate the propagation intensity function:

[0016]

[0017] Among them, tanh ensures that the propagation intensity is within the range of [-1, 1], 1 + |S i (t) - S j (t)| 2 prevents the intensity from infinitely expanding due to extreme sentiment differences;

[0018] S202: Consider the influence of the distance d of the sentiment propagation path ij on the propagation intensity, and introduce an exponential decay factor β represents the decay coefficient of the propagation path distance;

[0019] S203: By traversing all node pairs in the graph network, calculate the overall propagation intensity T p ;

[0020]

[0021] Among them, A ij represents the adjacency matrix value of nodes i and j in the graph network. The value is 0 or 1, indicating the presence or absence of a relationship. t 0 and t T represent the start point and end point of the time range of emotion propagation, and N represents the number of nodes in the graph network.

[0022] As a preferred solution of the dynamic network Internet emotion analysis method based on sample augmentation according to the present invention, wherein: the specific steps of the step three include:

[0023] S301: Introduce the time correlation function Ψ(t, C i ) to capture the propagation trend of node i over time;

[0024] S302: Combine the propagation intensity with the time correlation function to obtain the dynamic emotion trend, construct a prediction function, and predict the future emotion trend;

[0025] S303: Use an anomaly detection model to monitor abnormal behaviors in emotion propagation; the expression of the anomaly detection model is:

[0026] Among them, P k represents the propagation intensity of the k-th abnormal event, and R k represents the propagation range of the k-th abnormal event; M represents the number of abnormal emotion events; γ represents the number of abnormal emotion events, and T a represents the normalization adjustment of the overall impact on abnormal events.

[0027] As a preferred solution of the dynamic network Internet emotion analysis method based on sample augmentation according to the present invention, wherein: when T a is greater than the set threshold, it is regarded as abnormal propagation;

[0028] For the behavior regarded as abnormal propagation, if P k is significantly higher than the normal value, the abnormal propagation is identified as an extreme emotion outbreak. If the range R k abnormally expands, the abnormal propagation is identified as abnormal propagation.

[0029] As a preferred solution of the dynamic network Internet emotion analysis method based on sample augmentation according to the present invention, wherein: the calculation formula of the correlation function Ψ(t, C i ) is:

[0030]

[0031] Among them, w is the time period factor, Ci Represents the emotional context features of the nodes.

[0032] As a preferred embodiment of the method for analyzing Internet emotions in a dynamic network based on sample augmentation according to the present invention, wherein: the expression of the prediction function is:

[0033]

[0034] Wherein, T t The larger the value, the stronger the future emotional propagation trend.

[0035] As a preferred embodiment of the method for analyzing Internet emotions in a dynamic network based on sample augmentation according to the present invention, wherein: the predicted emotional propagation trend T t is combined with the detected abnormal propagation intensity T a to generate a comprehensive propagation evaluation value F, F = T t / T a ;

[0036] If F > 1, the adjustment method for the sample augmentation strategy is: focus on augmenting the sample data of high - impact nodes;

[0037] If F < 1, the adjustment method for the sample augmentation strategy is: preferentially augment the sample data of nodes with less emotional fluctuation to improve data balance;

[0038] If F = 1, it indicates propagation balance and the current strategy is relatively stable.

[0039] An analysis system applied to the above - mentioned method for analyzing Internet emotions in a dynamic network based on sample augmentation, the system includes the following working modules:

[0040] A data collection module, responsible for collecting user emotional data in real - time from multiple channels such as Internet platforms, social media, forums, news websites, etc.; an emotional mapping module, responsible for mapping the collected emotional data into a graph network and constructing the association between nodes and edges;

[0041] An emotional propagation graph network construction module, responsible for constructing a dynamic graph network according to the data collection and emotional mapping results, and dynamically updating the propagation path and intensity; an emotional propagation intensity calculation module, responsible for calculating the emotional propagation intensity and propagation range of each node based on the graph network;

[0042] A time - series emotional change prediction module, responsible for predicting the future emotional change trend of a specific group based on the emotional propagation path and intensity data; an abnormal emotional fluctuation monitoring module, responsible for monitoring the abnormal emotional fluctuations during the emotional propagation process and detecting the abnormal outbreak of emotions in a timely manner;

[0043] And a feedback optimization and strategy adjustment module, which is responsible for adjusting the analysis strategy and sample expansion strategy of the emotion propagation path according to the results of emotion prediction and anomaly detection.

[0044] The present invention also discloses a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned dynamic network Internet emotion analysis method based on sample expansion are implemented.

[0045] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned dynamic network Internet emotion analysis method based on sample expansion are implemented.

[0046] Advantages of the present invention:

[0047] 1. By constructing an emotion propagation model based on time series prediction, the present invention can dynamically capture the intensity change and propagation range in the emotion propagation process, accurately analyze the time series characteristics of emotions, and avoid the limitations of traditional static models;

[0048] 2. By introducing an anomaly detection mechanism, the present invention can monitor extreme events in emotion fluctuations in real time, identify abnormal emotion propagation in the network, and timely discover and warn of potential risks such as public opinion crises and cyber violence;

[0049] 3. The present invention uses sample expansion technology to optimize the training data of emotion propagation, improve the generalization ability of the model, and further improve the accuracy of emotion prediction. At the same time, through the optimized analysis of the propagation path and intensity, the simulation effect of emotion propagation can be effectively improved, making the emotion trend prediction more reliable;

[0050] 4. By combining emotion trends and anomaly detection results, the present invention provides a comprehensive emotion propagation evaluation mechanism, which helps decision-makers make more accurate predictions and response measures. It has important application values in fields such as advertising marketing, public opinion management, and social security. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings. Among them:

[0052] Figure 1 It is a schematic flowchart of the dynamic network Internet emotion analysis method based on sample expansion proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification.

[0054] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0055] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that can be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that mutually excludes other embodiments.

[0056] Refer to Figure 1 , for an embodiment of the present invention, a dynamic network Internet sentiment analysis method and system based on sample augmentation are provided. This method includes the following steps:

[0057] Step 1: Construct an emotion propagation graph network. First, collect users' emotion data from the Internet and map the collected emotion data into a graph network;

[0058] Among them, the nodes i, j of the graph network represent users or events; the edge A ij represents the relationship between users; the initial weight S i of the node represents the intensity of the emotion expressed by the user;

[0059] Step 2: In the constructed graph network, further calculate the intensity of emotion propagation through an algorithm. Through the analysis of the propagation path and intensity, key influencing nodes can be screened out. The construction of the graph network can regard emotion propagation as a complex interaction process, where the relationship between users (or events) determines how emotions flow between these nodes. By calculating the intensity of emotion propagation between nodes, we can understand the spread range of emotions in the network and the influence of each node, thereby identifying those "key nodes" that play an important role in emotion propagation. This step provides a basis for subsequent prediction and optimization.

[0060] The steps of calculating the intensity and range of emotion propagation specifically include:

[0061] S201: According to the difference in emotion intensity ΔS ij (t) = S i (t) - S j (t) between nodes i and j, calculate the propagation intensity function:

[0062]

[0063] Among them, tanh ensures that the propagation intensity is within the range of [-1, 1], and 1 + |S i (t) - S j (t)| 2 prevents the intensity from infinitely expanding due to extreme emotion differences;

[0064] S202: Consider the influence of the distance d of the emotion propagation path ij on the propagation intensity, and introduce an exponential decay factor β represents the decay coefficient of the propagation path distance;

[0065] S203: By traversing all node pairs in the graph network, calculate the overall propagation intensity T p ;

[0066]

[0067] Among them, A ij represents the adjacency matrix value of nodes i and j in the graph network, and the value is 0 or 1, indicating whether there is a relationship. t 0 and t T represent the start point and end point of the time range of emotion propagation, and N represents the number of nodes in the graph network.

[0068] Step 3: Utilize the propagation path and intensity data obtained in Step 2 to predict the future emotion change trend of a specific group by establishing the temporal relationship of emotion propagation. When predicting the emotion change trend, monitor abnormal emotion fluctuations in real time. Through in-depth analysis of the emotion propagation path and node characteristics, the fluctuations of group emotions can be predicted, and the emotion fluctuations that may trigger a crisis (such as the spread of sudden negative emotions) can be discovered in a timely manner.

[0069] Specifically, S301: Introduce a time correlation function Ψ(t, C i ) to capture the propagation trend of node i over time;

[0070] The calculation formula of the correlation function Ψ(t, C i ) is:

[0071]

[0072] Among them, w is the time period factor, and C i represents the emotion context feature of the node.

[0073] S302: Combine the propagation intensity with the time correlation function to obtain a dynamic emotion trend, construct a prediction function to predict the future emotion trend, and the expression of the prediction function is:

[0074]

[0075] Among them, T t The larger the value, the stronger the future emotional propagation trend.

[0076] S303: Use the anomaly detection model to monitor abnormal behaviors in emotional propagation; the expression of its anomaly detection model is:

[0077] Among them, P k represents the propagation intensity of the k-th abnormal event, and R k represents the propagation range of the k-th abnormal event; M represents the number of abnormal emotional events; γ represents the number of abnormal emotional events, and T a represents the normalization adjustment of the overall impact of abnormal events.

[0078] Among them, T a When it is greater than the set threshold, it is regarded as abnormal propagation. For the behavior regarded as abnormal propagation, if P k The abnormal propagation significantly higher than the normal value is identified as an extreme emotional outburst. If the range R k The abnormal propagation with an abnormal expansion is identified as abnormal propagation.

[0079] Step Four: Feed the results of emotional trend prediction and anomaly detection back into the emotional propagation graph network to further optimize the sample augmentation strategy. After analyzing and predicting emotional propagation, we need to optimize the model and augment data samples based on this information. By combining the emotional propagation trend with the anomaly detection results, we can obtain a feedback coefficient F), which represents the effectiveness of the emotional propagation pattern.

[0080] Combining the predicted emotional propagation trend T t with the detected abnormal propagation intensity T a to generate a comprehensive propagation evaluation value F, F = T t / T a ;

[0081] If F > 1, the adjustment method for the sample augmentation strategy is: focus on augmenting the sample data of high-impact nodes; if F < 1, the adjustment method for the sample augmentation strategy is: preferentially augment the sample data of nodes with less emotional fluctuations to improve data balance; if F = 1, it means propagation balance, and the current strategy is relatively stable. In practical applications, when F is approximately equal to 1, it can also be considered as propagation balance.

[0082] And the analysis system applied to the above-mentioned dynamic network Internet emotion analysis method based on sample augmentation. This system includes the following working modules:

[0083] The data collection module is responsible for collecting user sentiment data in real time from multiple channels such as Internet platforms, social media, forums, and news websites; the emotion mapping module is responsible for mapping the collected emotion data into a graph network and constructing the association between nodes and edges.

[0084] The emotion propagation graph network construction module is responsible for constructing a dynamic graph network based on the data collection and emotion mapping results, and dynamically updating the propagation path and intensity; the emotion propagation intensity calculation module is responsible for calculating the emotion propagation intensity and propagation range of each node based on the graph network.

[0085] The temporal emotion change prediction module is responsible for predicting the future emotion change trend of a specific group based on the emotion propagation path and intensity data; the abnormal emotion fluctuation monitoring module is responsible for monitoring the abnormal emotion fluctuations that occur during the emotion propagation process and detecting the abnormal outbreak of emotions in a timely manner;

[0086] And the feedback optimization and strategy adjustment module is responsible for feedback-adjusting the analysis strategy and sample expansion strategy of the emotion propagation path according to the results of emotion prediction and abnormal detection.

[0087] To further verify the beneficial effects of the present invention, the following experiment was conducted. Emotion data of 1000 users were obtained from a social media platform, including their emotion intensity, interaction information (such as comments, forwards), and timestamps, etc. The emotion data includes the emotion intensity expressed by users at specific time nodes. The text emotions were classified into positive, negative, and neutral using an emotion analysis algorithm, and their intensities were quantified numerically. A social network relationship among 1000 users (a graph network constructed based on friendship relationships and interaction information) was constructed, and the propagation intensity (emotion propagation path) of each edge was calculated.

[0088] The sample expansion strategy was optimized through the above method steps, and the effects of using the method of the present invention and the traditional static emotion analysis method were compared. The results are shown in the following table:

[0089] Table 1 is a comparison table of the prediction effects of emotion propagation trends

[0090]

[0091] Based on the above analysis, the prediction accuracy is as follows: It measures the degree of coincidence between the prediction model and the actual emotional change trend. Using the method of the present invention can capture the temporal changes of emotions more accurately, and the prediction accuracy is increased from 75% of the traditional method to 92%; the abnormal fluctuation detection accuracy: It refers to whether the model can effectively identify abnormal emotional fluctuations (such as sudden extreme emotions). The detection accuracy of the method of the present invention is 90%, which is much higher than 65% of the traditional method. Time consumption: It shows the time required for the model to run. Considering the sample augmentation technology of the present invention, more data can be processed in a relatively short time, and the time consumption is lower than that of the traditional method. Sample augmentation effect: It reflects the optimization effect of the sample augmentation strategy on the model training data. Using the sample augmentation method of the present invention increases the training data by 45%, greatly improving the generalization ability of the model.

[0092] This embodiment also provides a computer device, which is applicable to the situation of the dynamic network Internet emotion analysis method based on sample augmentation, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the dynamic network Internet emotion analysis method based on sample augmentation proposed in the above embodiment.

[0093] This computer device can be a terminal. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0094] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for internet sentiment analysis of a dynamic network based on sample augmentation proposed in the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM for short), electrically erasable programmable read-only memory (EEPROM for short), erasable programmable read-only memory (EPROM for short), programmable read-only memory (PROM for short), read-only memory (ROM for short), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0095] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A dynamic network Internet sentiment analysis method based on sample expansion, characterized in that: The method comprises the following steps: Step 1: Construct an emotion propagation graph network. First, collect users’ emotion data from the Internet and map the collected emotion data into a graph network. The nodes i and j of the graph network represent users or events; the edges represent the relationships between users; the initial weights of the nodes are S i Indicates the intensity of the emotion expressed by the user; Step 2: In the constructed graph network, the intensity of emotion propagation is further calculated through algorithms. Through propagation path and intensity analysis, key influencing nodes can be screened out; Step 3: Using the propagation path and intensity data obtained in step 2, the temporal relationship of emotion propagation is established to predict the future trend of emotion changes in a specific group. When predicting the trend of emotion changes, abnormal emotion fluctuations are monitored in real time. Step 4: Feedback the results of sentiment trend prediction and anomaly detection into the sentiment propagation graph network to further optimize the sample expansion strategy.

2. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 1 is characterized in that: In step 2, the step of calculating the intensity and range of emotion propagation specifically includes: S201: Based on the difference ΔS of the emotion intensity between nodes i and j ij (t) = S i (t)-S j (t), calculate the propagation intensity function: Among them, tanh ensures that the propagation strength is in the range of [-1, 1], 1+|S i (t)-S j (t)| 2 Prevent the intensity from expanding infinitely due to extreme emotional differences; S202: Consider the distance d of the emotion propagation path ij Impact on propagation intensity, introducing exponential decay factor β represents the attenuation coefficient of the propagation path distance; S203: Calculate the overall propagation strength T by traversing all node pairs in the graph network p ; Among them, A ij Represents the adjacency matrix value of nodes i and j in the graph network. The value is 0 or 1, indicating whether there is a relationship. t0 and t T represents the starting and ending points of the time range of sentiment propagation, and N represents the number of nodes in the graph network.

3. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 2 is characterized in that: The specific steps of step three include: S301: Introducing the time correlation function Ψ(t, C i ), used to capture the temporal propagation trend of node i; S302: combining the propagation intensity with the time correlation function to obtain a dynamic emotion trend, constructing a prediction function, and predicting future emotion trends; S303: Use the anomaly detection model to monitor abnormal behaviors in emotional communication; the expression of the anomaly detection model is: Among them, P k represents the propagation intensity of the kth abnormal event, R k represents the propagation range of the kth abnormal event; M represents the number of abnormal emotional events; γ represents the number of abnormal emotional events, T a Represents a normalized adjustment for the overall impact of an unusual event.

4. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 3 is characterized by: T a When it is greater than the set threshold, it is considered as abnormal propagation; For the behavior to be considered abnormal propagation, if P k Abnormal transmission that is significantly higher than the normal value is considered to be an extreme emotional outburst. Abnormal transmission with an abnormally expanded range Rk is considered to be abnormal transmission.

5. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 4 is characterized in that: The correlation function Ψ(t, C i ) is calculated as: Among them, w is the time period factor, C i Represents the emotional context features of a node.

6. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 5 is characterized by: The prediction function expression is: Among them, T t The larger the value, the stronger the trend of future sentiment propagation.

7. The dynamic network Internet sentiment analysis method based on sample expansion according to claim 6 is characterized by: The predicted sentiment propagation trend T t and the detected anomaly propagation intensity T a Combined to generate a comprehensive communication evaluation value F, F = T t / T a ; If F>1, the sample expansion strategy is adjusted as follows: focus on expanding the sample data of high-impact nodes; If F < 1, the sample expansion strategy is adjusted as follows: give priority to expanding the sample data of nodes with smaller sentiment fluctuations to improve data balance; If F=1, it means that the propagation is balanced and the current strategy is relatively stable.

8. The analysis system of the dynamic network Internet sentiment analysis method based on sample expansion according to claim 7 is characterized in that: The system includes the following working modules: The data collection module is responsible for collecting user sentiment data in real time from multiple channels such as Internet platforms, social media, forums, and news websites; The emotion mapping module is responsible for mapping the collected emotion data into the graph network and building the association between nodes and edges; The emotion propagation graph network construction module is responsible for building a dynamic graph network based on data collection and emotion mapping results, and dynamically updating the propagation path and intensity; The emotion propagation intensity calculation module is responsible for calculating the emotion propagation intensity and propagation range of each node based on the graph network; The time-series emotion change prediction module is responsible for predicting the future emotion change trend of a specific group based on the emotion propagation path and intensity data; The abnormal emotion fluctuation monitoring module is responsible for monitoring abnormal emotion fluctuations that occur during the emotion propagation process and timely detecting abnormal emotion outbreaks; As well as the feedback optimization and strategy adjustment module, it is responsible for providing feedback to adjust the analysis strategy and sample expansion strategy of the emotion propagation path based on the results of emotion prediction and anomaly detection.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the dynamic network Internet sentiment analysis method based on sample expansion according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the dynamic network Internet sentiment analysis method based on sample expansion according to any one of claims 1 to 7 are implemented.

Images